[FFmpeg-devel] Contract work
bostjan.strojan at gmail.com
Thu Feb 2 15:30:36 CET 2012
No really sure if i understand the requirements, but:
Flash players (like jwplayer) do usually support subtitling and/or logo
placement possibilities, so unless you really need for all this to be
hardcoded, that would seem like a good option. "They" also support html5
mode, which may not carry all the options from flash world thought, so i
would advertise that as: maybe later, separate contract.
On Sun, Jan 29, 2012 at 10:41 AM, Rene von Olten <renevonolten at gmail.com>wrote:
> Hello Developers,
> I am a freelance web designer based in France and for one of my
> clients (a non-commercial educational/game web site for a church
> organization) I'm looking for the following solution:
> They have videos hosted on Apache/linux server, 5 to 60 mins duration
> These videos need to be visitor personalized, each video contains a
> static scene showing a bulletin board or school board and the name of
> the visitor needs to be displayed on this school board
> Videos are encoded on local PC in h264 using Telestream Episode, they
> are flexible on this and can change to any other video encoder.
> Important is that they code the videos on local computers and then
> upload to their remote Apache server.
> When creating the customized video, to avoid long render times the
> existing videos should not be re-encoded, only the part with the
> bulletin board should be re-encoded and the final video segments need
> to be merged into one video file which can be downloaded by the user
> or viewed in browser flash or html5 video player.
> At first for development/testing purposes I will have 3 video segments
> available at my own server:
> 1) First part of video
> 2) Part of video which is static and contains the bulletin board.
> 3) Last part of video
> The following steps need to be applied to these segments:
> 1) (for example) generate a still image that consists of the user
> name, for example "David".
> 2) Merge this still image with the 2nd part of the video so that
> "David" appears on the bulletin board. This can be a simple merge, no
> need to worry about alpha channels, masks, etc. (maybe later), encode
> to new merged output. If you can skip the creating of the still image
> and merge "David" into the video file directly, great.
> 3) Merge the 3 parts of the video into one large file without
> re-encoding part 1 and 3
> At a later stage the splitting of the video into three sections before
> merging the visitor name should be done on the server as well, so
> there will be one stock video available, your solution splits this
> into three parts, re-renders part 2 to insert the name, and then
> merges the 3 parts into 1 video file again.
> Please let me know if you think you can create the above solution for
> us and an indication of your rate for this.
> René von Olten
> ffmpeg-devel mailing list
> ffmpeg-devel at ffmpeg.org
More information about the ffmpeg-devel