Re: Questions about VIDIOC_G_JPEGCOMP / VIDIOC_S_JPEGCOMP

2009-02-20 Thread Hans Verkuil
On Thursday 19 February 2009 15:02:31 Hans Verkuil wrote: Hi, The VIDIOC_G_JPEGCOMP / VIDIOC_S_JPEGCOMP v4l2 ioctls seem not to be used by many drivers / applications. They should! Unfortunately, these ioctls are completely undocumented. Which might be the reason why they aren't used

Re: Questions about VIDIOC_G_JPEGCOMP / VIDIOC_S_JPEGCOMP

2009-02-20 Thread Jean-Francois Moine
On Fri, 20 Feb 2009 09:29:36 +0100 Hans Verkuil hverk...@xs4all.nl wrote: Support for these ioctls should be added to v4l2-ctl.cpp. It's the right place for that. But more important is to document these ioctls in the v4l2 spec. As far as I can tell these ioctls came from the zoran

Re: Questions about VIDIOC_G_JPEGCOMP / VIDIOC_S_JPEGCOMP

2009-02-20 Thread Hans Verkuil
On Friday 20 February 2009 12:04:00 Jean-Francois Moine wrote: On Fri, 20 Feb 2009 09:29:36 +0100 Hans Verkuil hverk...@xs4all.nl wrote: Support for these ioctls should be added to v4l2-ctl.cpp. It's the right place for that. But more important is to document these ioctls in the

Re: Questions about VIDIOC_G_JPEGCOMP / VIDIOC_S_JPEGCOMP

2009-02-20 Thread Trent Piepho
On Fri, 20 Feb 2009, Jean-Francois Moine wrote: So, I propose to remove these ioctls, and to add two controls: one to set the JPEG quality (range 15..95 %) and the other to set a webcam quality which might be a boolean or any value depending on some associated webcam parameter. A control can

Re: Questions about VIDIOC_G_JPEGCOMP / VIDIOC_S_JPEGCOMP

2009-02-19 Thread Hans Verkuil
Hi, The VIDIOC_G_JPEGCOMP / VIDIOC_S_JPEGCOMP v4l2 ioctls seem not to be used by many drivers / applications. They should! Unfortunately, these ioctls are completely undocumented. Which might be the reason why they aren't used :-) In some ms-win traces, there are automatic and dynamic