From: SourceForge.net <no...@so...> - 2010-05-30 13:08:13
|
Patches item #3007214, was opened at 2010-05-26 07:34 Message generated for change (Comment added) made by paulrho You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=302341&aid=3007214&group_id=2341 Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: None Group: None Status: Open Resolution: None Priority: 5 Private: No Submitted By: Paul S (paulrho) Assigned to: Roy Rankin (roy_r_rankin) Summary: Re-implement PAL Video Module Initial Comment: Currently the PAL Video Module is disabled (commented out), one reason is the internal methods it used are not supported in current code. I have recoded this so that it works. (Once submitted/provided, I would like to have someone check that I have done this in a way that makes sense). Also I need to check that I'm not consuming too much CPU in the way I've done the mod. I would also like comments on the actual use of IN1 and IN2 inputs to read the PAL signal. I have used these signals as SYNC and LUMINANCE. In a true PAL composite signal, these are superimposed such that : (not using actual values, but example values) sync is 0 no sync black is 0.3 (say) no sync white is 1.0 I think that in the very least the two pins should be labelled SYNC and LUME, or possibly move to the composite one input (analog) method. Retaining two pins, SYNC and LUME, makes it easier from a wiring point of view as the resistors required do not need to be implemented on the virtual breadboard. (And it then simply makes the change as simple as a rename of pins for clarity) The proposed change involves the following: * Recode where necessary the module to use the new methods * Rename in1 and in2 to be SYNC and LUME * SYNC is basically either ON or OFF (low means SYNC) * LUME then becomes an analog light level with 0=black 5v=white - anywhere inbetween a shade of grey (proportional) * Include a simple PAL test program to examples In my example I've only used LUME either on (white) or off (black). Opened as a tracker artifact (I hope I'm doing the right thing), will attach svn diffs shortly. Paul S. ---------------------------------------------------------------------- >Comment By: Paul S (paulrho) Date: 2010-05-30 13:08 Message: With the above changes in the trunk: I've added another svn diff on top of that (see attached) to speed up the render (suggested by Ralf F): Rather than replot every pixel, only those that change are rewritten to the image. As the speed increase is out-of-sight, I've also increased the x resolution to double, to make it similar to the appropriate ratio. I can now get near realtime display of the example I'm using. Only downside would be if there are actual problems in the video signal, it will leave a persistent image (although that is probably not a bad thing). Thanks Ralf! ---------------------------------------------------------------------- Comment By: Roy Rankin (roy_r_rankin) Date: 2010-05-27 11:31 Message: patch committed to SVN, but should be add back into documentation. ---------------------------------------------------------------------- Comment By: Paul S (paulrho) Date: 2010-05-26 13:04 Message: Added a simple (early version) of a PIC program that outputs (two pin) PAL video. place both in a directory and run with : $ gpsim -c paltv-stim.stc You'll note: (1) it runs really slowly (about 1/60th of realtime) (2) input on the left is stimulated automatically, and the other stays fixed (not stimulated) ---------------------------------------------------------------------- Comment By: Paul S (paulrho) Date: 2010-05-26 12:52 Message: attached file with svn diff of the modules directory (all changes in there) ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=302341&aid=3007214&group_id=2341 |