On Friday 19 April 2002 09:57 am, Munaut Sylvain wrote: > Hi > > My first question is : Are all usb webcam using about the same protocol > ( i mean is there a lot of similarities between protocols or are they > all completel different ? )
Never twice the same protocol it seems (okay, there may be some, but...) > And the second one : I suppose some people have already done this ( > guessing protocols ). Do you have some tip on how to do ? On the sourceforge pages for the 3com homeconnect (http://homeconnectusb.sourceforge.net) we've chronicled the details of our reverse engineering efforts. We too had a camera whose creators refused to release details and had to start from scratch (in 3com's defense, they did not create the camera and did not have a license which allowed them to give us access to the specs we needed). > I use a USB spy program under win32 to listen to the USB port while i'm > using the web cam. Do you knwo better spy program than "usb snoop" ? usb snoop is the tool I used to do the reverse engineering on the homeconnect. You can read what we've done, you generally start by viewing the captured data as gray scale, this gives you an idea of whether or not you're looking at planar RGB, YUV, or some other format. Then you start pointing the camera at sheets of your daughter's colored construction paper, break out your spreadsheet and start doing statistical analysis on the raw data against a "good" end product from the Win32 driver (what does the image payload data look like for an almost entirely red image?) Takes a lot of perserverance, but it can be done. Honestly I found navigating the pitfalls of a v4l driver more difficult than reverse engineering the camera. _______________________________________________ [EMAIL PROTECTED] To unsubscribe, use the last form field at: https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
