Date: Wed, 18 Jun 1997 21:38:47 -0400 From: Randall Hopper <rhh@ct.picker.com> To: multimedia@freebsd.org Subject: Fxtv 0.43 Message-ID: <19970618213847.33721@ct.picker.com>
next in thread | raw e-mail | index | archive | help
URL: http://multiverse.com/~rhh/fxtv Got some new toys put into this version. - Image Save-to-Disk - saves the last frozen image (TIFF) - Audio Save-to-disk (Raw, AU, AIFF, WAV, VOC, MPEG-2, MPEG-3) - Video Save-to-disk (TIFF sequences and/or MPEG video stream) - FPS control added - Doesn't kill the CPU anymore. Virtually no load for direct video mode, and surprisingly little load even in ximage mode. And a few things (see the ChangeLog for details). Be sure to check out the sample clips, and keep in mind the 30fps clip was captured on an IDE disk if you can believe it, so you SCSI folks will be doing even higher res at the same rate. A few notes about the new features. There are some new build- and run-time dependencies (libtiff, sox, mpeg_encode, etc.). See the README for details on building, installing, and running this version and the utilities it calls. Also, 0.43 delivers on a slightly hacked version of the bt848 driver. FPS adjustment didn't work with previous versions, but with this version it at least works for single-field. The Bt848's FPS control is used both for throttling the video stream while capturing to disk as well as for displaying continuous video on the screen. As always, feel free to add things into fxtv and mail new features/ comments/bug reports/etc. anytime. There's plenty of room for video capture optimization and support enhancements (what's there now is really just first-cut). A few of the many things that still can be done to to beef it up include: - interleaved capturing to multiple disks - simultaneous audio and video capturing - capturing 24-bit instead of 16-bit if there's enough disk bandwidth available - capturing YUV (instead of RGB and converting later), and - making the "Optimize" algorithm more intelligent :-) It'll also be worth trying capture to an LFS once John Dyson gets it put back together (seems like I read that he's working on it). Should give better throughput. Incidentally, I did try capturing to raw hard disk devices and found I got a higher consistent FPS throughput (i.e. no "hickups") writing to a UFS, so I didn't put in support for capturing to device files in 0.43. Easy enough to hack in though if somebody wants to play with it. Have fun, and let me know how it goes! Randall
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?19970618213847.33721>