Hi! > If you haven't merged the patch yet, don't bother.
Already done but not committed in the SVN repo. > There's a lot of redundant stuff in there which I am cleaning up right > now. Also, thanks to Francks suggestion to investigate the m5602.csr > file I'm able to piece together on how to setup the chip properly. Would you send me over the final patch once you finish cleaning up the code? > I'm currently idling in the #m560x channel on FreeNode, drop by anytime > if you have the time. I can do that during the week-ends. I spend the week (working days) in another town for work, where I have limited Internet access. Erik, I have another task for you :) I would like you to investigate the standard RGB output from the cam, instead of the raw bayer stream. The ov9650 is able to output such a normalized format, i.e RGB 16bit (555 or 565) and it would be wonderful for the driver to retrieve such a format directly from the cam so that applications can use it without doing any expensive software format conversion. I think you'll have to deal with registers 0x12 and 0x40 of the ov9650 sensor (see the datasheet, check google for that). What do you think about this? Best regards, Ilyes Gouta. > Regards, > Erik > > > Ilyes Gouta wrote: > > Hi! > > > > Thanks Erik! I'm going to review it and merge it. This is the kind of > > help we're expecting in order to move forward with this driver! > > > > As far as the pitch is concerned, I'm actually looking if there is > > some register (it should be) in the ov9650 (or the bridge) that the > > driver should write to, to instruct the cam to return the right > > picture width. For example, on the mt9v011 sensor, such a > > functionality is provided by register 0x03. I think that pitch > > variation should be handled by the hardware (the cam) and *not* by the > > software, i.e we don't get a picture from the cam then apply a > > transformation using a software routine before displaying it. The > > current workaround in LiveCam, that I use to display a proper picture, > > given a display area of w pixels (let's say 640) and an unknown > > picture pitch x (greater than W), is that it skips x - w bytes from > > the bayer stream once the decoder reads w bytes/pixels. You can then > > apply a bayer decoder to recover the missing color components using > > interpolation. That's it! :) > > > > Thanks again! > > > > Best regards, > > Ilyes Gouta. > > > > 2008/2/10 Erik Andrén <[EMAIL PROTECTED]>: > >> Hi, > >> By sniffing the windows driver I have created the attached patch which > >> inits > >> the webcam some more. > >> Using this patch, the camera works reliably every time for me, not needing > >> to reinsert the module a number of times. > >> I refactored the code a bit and pulled out the actual hex data into a > >> couple of arrays and used some loops to create the actual writes, as I felt > >> that this made the code more readable and separated the actual firmware > >> data > >> from the initialization function. I'm not too familiar with svn but created > >> the patch with using "svn diff", I hope this is the right way of doing it. > >> > >> Also, I've tried to investigate the pitch problem but haven't fully wrapped > >> my head around the problem. > >> What is needed to be done with a frame and the pitch to get a proper image? > >> > >> Please review the patch and commit if appropriate! > >> > >> Regards, > >> Erik Andrén > >> > >> > >> > >> ------------------------------------------------------------------------- > >> This SF.net email is sponsored by: Microsoft > >> Defy all challenges. Microsoft(R) Visual Studio 2008. > >> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ > >> _______________________________________________ > >> M560x-driver-devel mailing list > >> [email protected] > >> https://lists.sourceforge.net/lists/listinfo/m560x-driver-devel > >> > >> > > > ------------------------------------------------------------------------- This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ _______________________________________________ M560x-driver-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/m560x-driver-devel
