| Commit | Date | |
|---|---|---|
|
[r189]
by
bruno.herbelin
Yet again BUG FIX ; error in stop introduced by some recent changes, cleaning of code (drop option to use a thread for video refresh ; the idea is bad as it uses signal/slots to inform video source to update frame, which is not very robust between threads) |
2010-07-29 22:37:33 | Tree |
| 2010-07-29 19:52:04 | Tree | |
| 2010-07-29 15:56:27 | Tree | |
| 2010-07-28 21:49:21 | Tree | |
| 2010-07-28 21:44:28 | Tree | |
|
[r184]
by
bruno.herbelin
More BUG FIX in Video FIle : this time, I think its much better : 1) I do nothing special in the refresh timer (so there is no jump/ jerk of picture) BUT, 2) when the refresh timer start to be late on presentation time, it tells the Decoding thread to skip the next frame. This way, 1 or 2 frames later, the presentation time stamp to show is 1 frame later, so we can catch up. (+ compilation complain fixed in Rendering Manager XML). |
2010-07-28 21:29:10 | Tree |
| 2010-07-27 11:19:33 | Tree | |
| 2010-07-27 11:04:09 | Tree | |
|
[r181]
by
bruno.herbelin
AAAARG ! Fine-tuning and debugging of VideoFile threads is horrible... This time, it looks like its working fine. |
2010-07-27 10:45:32 | Tree |
|
[r180]
by
bruno.herbelin
DISABLED refresh timer THREAD : this does not improves anything, and adds yet another thread to synchronise. BUT, some important BUG FIX in time management (e.g. start at begin, not 0), and implementation of FRAME SKIPPING mechanism when the delay between presentation time stamp and timer becomes too big. |
2010-07-27 00:48:20 | Tree |