Cameron Gutman
|
12e34a9f4e
|
Add the code back to request an IDR frame when switching back to Moonlight
|
2016-07-10 22:23:01 -07:00 |
|
Cameron Gutman
|
f243685990
|
Allow packet receiving to happen in parallel with video decoding by not using direct-submit
|
2016-06-09 13:33:31 -05:00 |
|
Cameron Gutman
|
69ba5a709b
|
Flushing the decoder seems to hang when the software fallback is in use, so let the destructor abort callbacks instead
|
2016-06-09 13:25:18 -05:00 |
|
Cameron Gutman
|
6c13c9b67d
|
Only show the NaCl plugin when a frame has been rendered
|
2016-06-09 13:22:27 -05:00 |
|
Cameron Gutman
|
1f3f4bd0aa
|
Implement fallback to software when hardware decoding is unavailable
|
2016-06-09 12:23:23 -05:00 |
|
Cameron Gutman
|
d08d1f3132
|
Fix the GL error assert macro
|
2016-05-10 01:33:04 -04:00 |
|
Cameron Gutman
|
64d9f18c39
|
Unbind the graphics context when exiting the stream
|
2016-05-09 22:54:07 -04:00 |
|
Cameron Gutman
|
0d8f82a5f9
|
Revert "Clear the display when tearing down the decoder. This closes #50"
This reverts commit 68aecaeed5139c8d6ad4f4a0cda4b60bd3c1b94f.
|
2016-05-09 22:35:12 -04:00 |
|
Cameron Gutman
|
68aecaeed5
|
Clear the display when tearing down the decoder. This closes #50
|
2016-05-09 17:02:16 -04:00 |
|
Cameron Gutman
|
f0d974bb73
|
Add profiling code to the video renderer
|
2016-05-09 14:26:23 -04:00 |
|
Cameron Gutman
|
8817df3f58
|
Use direct submit mode for better performance at higher CPU load
|
2016-05-05 17:09:49 -04:00 |
|
Cameron Gutman
|
33a19a52f4
|
Use the same context attributes as the video_decode sample
|
2016-05-05 17:03:05 -04:00 |
|
Cameron Gutman
|
21dd55566e
|
Decouple renderer from decoder and remove associated workarounds
|
2016-05-05 16:57:26 -04:00 |
|
Cameron Gutman
|
392c6f140a
|
Clear the shaders after destroying them. This closes #22
|
2016-03-13 19:08:12 -05:00 |
|
Cameron Gutman
|
47ede93af0
|
Update common submodule
|
2016-03-03 21:15:48 -05:00 |
|
Cameron Gutman
|
abe66c7e21
|
Set some nicer Graphics3D context attributes to allow the GLES backend more flexibility when choosing a context
|
2016-02-26 17:20:10 -05:00 |
|
Cameron Gutman
|
e7b45e5b6c
|
Set the graphics context size using the width and height of the stream instead of the initial size of the NaCl element. This closes issue #16
|
2016-02-25 22:48:22 -05:00 |
|
Cameron Gutman
|
24cf97b0ae
|
Only draw frames that are newer than the last frame we drew
|
2016-02-24 13:55:35 -05:00 |
|
Cameron Gutman
|
36e76ed86d
|
Use h264bitstream library to mixup the SPS so OS X will use hardware decoding
|
2016-02-21 15:57:45 -05:00 |
|
Cameron Gutman
|
83ba8a10bb
|
Force hardware acceleration for video decoding so we can see clearly when it's not working
|
2016-02-20 22:54:16 -05:00 |
|
Cameron Gutman
|
88dc8623cd
|
Send the SPS, PPS, and I-frame together to fix VideoToolbox hardware decoding on OS X. This closes #7
|
2016-02-20 22:28:58 -05:00 |
|
Cameron Gutman
|
5d25874131
|
Clear the decode queue on mouse lock or focus gain. This closes #5
|
2016-02-18 13:25:49 -05:00 |
|
Cameron Gutman
|
cf5b2ccbb4
|
Flush the decoder before deleting it
|
2016-02-18 12:04:17 -05:00 |
|
Cameron Gutman
|
441047db9b
|
Use 4 slices per frame for increased software decoding performance
|
2016-02-18 02:18:59 -05:00 |
|
Cameron Gutman
|
f8fb94f5e2
|
Optimize the renderer by keeping state changes to a minimum, more explicitly initializing the GL context, and calling glClear once per frame
|
2016-02-17 00:32:18 -05:00 |
|
Cameron Gutman
|
fe0120ca16
|
Allow software decoding fallback for OS X
|
2016-02-16 12:33:33 -05:00 |
|
Cameron Gutman
|
f77979d904
|
Increase rendering performance significantly
|
2016-02-13 15:48:33 -05:00 |
|
Cameron Gutman
|
b7032a8957
|
Rendering is now working
|
2016-02-13 14:49:27 -05:00 |
|
Cameron Gutman
|
e67a841a3e
|
Rendering WIP
|
2016-02-13 12:14:13 -05:00 |
|
Cameron Gutman
|
9a38318210
|
Video decoding is partially working
|
2016-02-13 09:12:41 -05:00 |
|