Hi, On Friday 14 November 2008, [EMAIL PROTECTED] wrote: > 2008/11/12 Laurent Pinchart <[EMAIL PROTECTED]>: > > On Tuesday 11 November 2008, [EMAIL PROTECTED] wrote: > >> Dear Laurent and other developers of linux-uvc: > >> > >> I was talking with my mentor. We are interesting in: > >> · Still Image capture: method 2 and method 3, but we don't know which > >> camera can use method 3. > > > > Here's a non exhaustive list of webcams that support (or at least claim > > to support) still capture method 2. > > > > 0402:5606 ALi Corp. > > 0408:030c Quanta Computer, Inc. > > 046d:0994 Logitech, Inc. > > 046d:09a1 Logitech, Inc. > > 046d:09b2 Logitech, Inc. > > 0471:0332 Philips > > 04f2:b016 Chicony Electronics Co., Ltd > > 05ac:8502 Apple Computer, Inc. > > 064e:a110 Suyin Corp. > > 0ac8:332d Z-Star Microelectronics Corp. > > 0ac8:3343 Z-Star Microelectronics Corp. > > 0ad2:933d Service & Quality Technology Co., Ltd > > 0c45:62e0 Microdia > > 0c45:63e0 Microdia > > 145f:013e Unknown > > 145f:013f > > 174f:5931 Syntek > > 174f:5a31 > > 174f:8a33 > > 1871:0306 > > 1c4f:3000 > > 1cac:a332 > > 1cac:b288 > > 5986:0104 Bison > > 5986:0200 > > 5986:0202 Bison > > 5986:0240 Bison > > 5986:0300 > > 5986:0303 Bison > > I don't know of any camera supporting still image capture method 3. > > I want to know what is method 0. I think that it's when webcam hasn't > any button => no interruption point. Is it true?
Not really. Snapshot functionality is not related to hardware triggers. The host can decide to take a snapshot at any time. All devices "support" method 1 even if they claim not to, as that method doesn't require any support from the device. > I will use this list for buy a good webcam for do this work. How can I > be the developer of "Still Image" methods? The most interesting for me > is the method 2. Just implement the function and submit a patch. To make sure your patch will be accepted you should submit intermediate versions and discuss implementation details on the linux-uvc-devel mailing list. > >> · Clock: I can see an example in documentation of UVC 1.1 (Class and > >> FAQ) with clock reference, but it's only a unique (and duplicated) > >> example with stream payload. > > > > See section 2.4.3.4 in the UVC 1.1 specification. > > I'll read this section in this weekend. > > >> · UVC stub: My mentor is thinking about have a software UVC (like > >> mac80211 for wifi devices) that show you usb trams. It can be > >> interesting for learning and for debugging. > > > > I'm not sure to understand. Could you please elaborate ? > > My mentor is interesting in get a "software" USB device for get the > information of USB trams (USB controller <-> USB device). Like > mac80211 Test Driver ( > http://kernelnewbies.org/KernelProjects/Mac80211TestDriver ) but for > UVC. I'm not sure that would really be interesting. The purpose of the 802.11 test driver is to provide a virtual base to test the whole 802.11 stack (kernel space and user space). There's already an available V4L2 test driver (called vivi) that can be used for userspace testing. It would be much more interesting to implement a UVC gadget driver. See this thread for instance http://lists.berlios.de/pipermail/linux-uvc-devel/2006-September/000816.html. Best regards, Laurent Pinchart _______________________________________________ Linux-uvc-devel mailing list [email protected] https://lists.berlios.de/mailman/listinfo/linux-uvc-devel
