On Monday 27 August 2007 16:49:17 Pigeon wrote: > > I will try your patch but it may take a few days. Can I use the patch > > with 0.9.11-rc1? > > I've just reorganized my patch now, it's now split into two: > > http://pigeond.net/~pigeon/fgfs.plib.ss.20070828.patch > > http://pigeond.net/~pigeon/fgfs.osg.ss.20070828.patch > > > The plib one should work with 0.9.11-rc1
I was able to get the old patch (with some minor tweaks) to work with 0.9.11-pre1. I also did some testing. This would be a real nice addition to FlightGear and I would like to see it become part of the next release. > > > One of the things that I think is useful about the way I am doing this > > with yukon is that I can focus on flying during the FDM capture session > > without having to think about things like camera angels and zoom. Then > > when playing back the session during capture I can forget about the > > actual flying and focus on things related to the composition (camera > > angle, zoom...). In other words it cleanly separates the pilot and > > camera man functions. Of course this should be possible with your patch > > as well. > > Indeed. Your approach is good in many ways, and it should work with > any mean of capturing. There are both advantages and disadvantages: Pluses. 1. Better control over composition while recording. 2. No distractions related to the recording process while piloting the aircraft. 3. Since you can run the record playback session at lower resolutions than you would normally run FG while flying you can crank up the OpenGL quality parameters while recording. Minuses 1. Sounds do not playback very well using this method. The engine/prop sound does not change like it does normally. Wind noises from lowering the landing gear are not present. 2. At least with the PC-7 none of the engine gages work while using an external FDM. Perhaps this is related to #1. Is this a PC-7 issue or does this happen with all aircraft? 3. Does not allow you to record other aircraft that may have been present during the session when you recorded the FDM. So this will not work if you want to create a video of a formation flight for example. 4. During FDM play back movements of the controls are jerky and do not completely reflect what happened when the FDM was recorded. Hopefully #1, #2 and #4 will be fixed at some point. > > > Question - when your patch > > is doing the capture does it also capture the cursor and menu? > > At the moment, the plib version captures menus and dialogs, but not > the cursor. The current osg version, which I've used render-to-texture > to capture, only captures what's in the scene graph, which means it does > not capture the menu. In theory I could also implement a RTT version for > the plib branch as well. With yukon it is fairly simple to avoid recording the menus since you can tell it to ignore parts of the application window it is recording. But I think that your code should be designed to prevent things like menus from being captured and it looks like you are already there with OSG. Not recording the cursor is a good feature and a definite plus compared to yukon. > > > There is no sound since the version of yukon I was using does > > not capture sound (it will be in the next release). > > That would be an advantage of (future) yukon to capture audio as > well. Of course I could also implement that if it is worthwhile. > Sound would be a plus but I don't think it is a big deal at this point. Focus on getting video capture working as well as possible before worrying about sound. > > I've also tested the streamer under windows and it works pretty well > too. I believe I would not be straight forward to get yukon on windows, > but then I don't know. I believe that yukon is X11 and x86 or x86_64 only at this point. I would expect that we will see it for other architectures on X11 before we see it on Windows. But I believe that there are currently other viable options for Windows for those who want to capture videos of games and sims. For X11 users I get the impression that yukon is about the only viable option. At least that is what users on the gentoo forums are saying. > > The original intention of my streamer is to provide an easy and in > theory efficient way to capture video of FG, comparing to using any > external on-screen-capture tool. I know a few people was using FG's > jpg-httpd for capturing and making video, which is IMHO insanely > inefficient. > > I have not tried yukon myself yet, but with the way it works it > should be very fast as well. Yes yukon is fast and it has fairly low CPU overhead. It is actually a good general purpose OpenGL video capture tool considering how young it is. But at higher resolutions it starts to become IO bound because of the size of it's output files. It also needs a huge amount of disk space - a 640x480 capture uses about a half gigabyte per minute of recording. The advantage however is that you can do things like two pass processing, for better quality at a given compression level, that would be at best difficult using your patched FG code. Another plus for yukon is that you can start/stop recording by pressing F8 (this is the default key but it can be changed). For example, you may want to only record a few segments of a flight and yukon will let you do that. One other thing that may be an issue. I like to run FG in game mode across both of my monitors. This means that if I wanted to capture a live session (IE. not using a recorded external FDM) there would be no way to start mencoder after FG was running. And as you know starting mencoder before FG is running does not work. I hope this was helpful. Hal ------------------------------------------------------------------------- This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now >> http://get.splunk.com/ _______________________________________________ Flightgear-users mailing list Flightgear-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-users