The VideoLAN Forums

Hi All,

I'm working for a company that is building several test vans that have multiple video feeds from multiple vans. We are looking for a modified VLC re-encapsulation DLL that we can call with a few parameters that will allow us to stream an MP4 video/audio stream being transported by MPEG2-TS to a MP4 video file playable by VLC/WMP/MPC and editable by Ulead.

We are able to post process the stored MPEG2-TS packets from file through VLC to create an MP4 file, but we would like to do this as the streams are being received. We also want some control over the DLL to start/stop the capture. We would need to capture up to 8 streams from 8 seperate IP Addresses at once and more in the future.

We are seeking a professional group that is willing to develop this library. As far as we are concerned the DLL can remain open source and we will be able to fund the development. Our time frame for a complete DLL would be about 4 to 6 weeks.

Please PM me and I will be happy to discuss the requirements.

Thank you

Is this task too difficult? Am I too far off base with what we need? I'm open for suggestions too.

Thanks

What is the expected quality and resolution of the picture? Does the picture require sound? Describe end result in more detail if possible.

Can you provide a sample of the source you will be working with? Is the source formats and container consistent (always the same)? Is the source resolution consistent?

When you say that the end result must play on WMP, MPC and VLC. The only format that is common here (to all players) is MPEG43 and it does not work in a MP4 container.

The formats that work reliably with an MP4 container are MP4v (as found in Sorenson or 3ivx) or h.264. These are not native to Direct Show (WMP or MPC).

These are just some of the issues that would cause you to not get an answer.

The picture quality will be standard NTSC quality from live camera feeds monitoring tests. We hope to be able to provide HD quality in the future, but for now NTSC is fine.

I can provide source formats. In both PS and TS encapsulation.

As for the format, I will talk to our customer to find out what is acceptable.

Thank you for the response. We were trying to figure out why no one wanted to offer help or post a message.

--Brian

The picture quality will be standard NTSC quality from live camera feeds monitoring tests. We hope to be able to provide HD quality in the future, but for now NTSC is fine.

Is this to be a simultaneous feed to a satellite?

I'm assuming a multi-camera shoot where the cameras are feed to a switcher for angle, segways and effects, is this correct? If it is, the Make and Model of the Cameras and Switcher would be appreciated so that specs can be gained to see the range of possibilities for your particular application.

Also, it would be nice to know how you plan to handle the sound? As a Musical Event would require a separate truck and this can create sync issues and a few other technical issues. If Musical events are not to be included, then can it be assumed that the sound will be handled through a small mixer or the switcher itself and shotgun Mikes will be used (as in a Sound Stage)? Or do you plan to use Mikes attached to each Camera feed to the Switcher? The last question on this topic is, if the sound will always be Mono or is stereo a requirement? I'm trying to see concept here, so feel free to add on if possible.

So, can MPEG 2, interlaced, 320x240 Video @ 29.97 FPS with MPEG 2 audio be assumed?

I can provide source formats. In both PS and TS encapsulation.

Yes! Please do! This could answer lots of questions. A time limited link to the samples should work for evaluation purposes. Assuming these are more than test screens it could answer some of the questions already asked here and / or generate some other questions.

As for the format, I will talk to our customer to find out what is acceptable.

Ah! Yes! This is always the big question of quality versus resolution for what is reliably stream-able across an Internet connection and Today this should include handheld devices.

I will try to make some suggestions as I can get a few more answers from you.

Thank you for the response. We were trying to figure out why no one wanted to offer help or post a message.

--Brian

Your welcome! At least this should provide some basis to open a dialog for anyone who is interested to join in.

Okay,

This is nothing that complicated. It in fact is a very simple system. We have 8 Standard definition cameras that track a target as it is being tested. The 8 cameras are feed into 8 encoder boxes that encode the video stream to MPEG4 and transmit it via MPEG-TS. The IP Streams are broadcast on a closed network that can be tuned into and displayed on any of many displays.

As for the audio, we have one channel that will be recorded on all MPEG broadcasts.

This is one system, and there will be other systems exactly alike. Each system can be linked to the other so that any video broadcast can be received by any other system's receivers.

I will state this again. We have already captured the IP stream to disk (stripping off the RTP headers) and post process the video using VLC to create a playable video. We would like to hire a programming group to create a DLL that we can interact with from our software that can do the same. We would specify the IP address and the output file and tell the DLL to start and stop. The DLL (or 8 instances of the DLL) need only be able to record the 8 streams from the local system, not any other system as they will have the same software.

I will have sample video streams shortly that represent the exact video broadcast formats we will be using,

Thanks, i hope this helps.

Brian