WTV is the standard format Windows Media Center writes recording on.
We need a small program that read a WTV file and writes timing info and a hex dump of each component as the file advances. Audio and video is to be skipped - it's everything else (that would be for example the program guide) that we need, and only an hex dump, i.e. you don't need to parse the data, just extract it.
The format is not officially documented but there's some good docs, including:
[login to view URL]
Most importantly, ffmpeg supports it decently.
Specific things we need:
- Your program must be C/C++ only and it must read the file on its own. You CAN (and probably should) use ffmpeg and any portable C library. You cannot use DirectShow or any other Win only library.
- Sometimes WTV files are encrypted. However, their metadata is not (for example, you can still retrieve the program name or the station name from an encrypted file, even if you cannot decode the video or the audio). It's important that you are able to handle these files.
Additionally:
- Your work will be released as GPL. Feel free to add credits if you want, up to you.
- Because of the previous line, everything you use must be GPL compatible, please don't use proprietary libraries or code.
Budget:
- Please don't assume that because we're accepting bids up to $799 it's OK to just bid $775. Take some time to analyze the requirements and bid accordingly. Don't overbid. Try not to underbid (even though if you underbid and then as the project progresses it becomes obvious you need to increase the budget we will accept).
Hi.
If ffmpeg supports WTV container well - then it should be not a big deal to parse and extract the data you need even without lot of documentation (I hope). I was working with settopboxes before and have some knowledge about transport streams which contain different pids with video, audio, subtitles, EPG and so on.
So I could promise - this will be done in C/C++ with some ffmpeg usage or parts of it needed for parsing that type of container. So what do you think?
Regards.
Looks like most of the parsing work has been figured out in ffmpeg, I'd take the bits from ffmpeg that do the parsing add the hex writing code and add create a stand alone program out of them.