Search Issue Tracker
Fixed in 2017.3.0f3
Fixed in 2017.2.X
Won't Fix in 2017.1.X
Votes
9
Found in
5.6.2p1
Issue ID
926390
Regression
No
VideoPlayer.Stop() causes a performance spike
To reproduce:
1. Open project attached by QA
2. Open "x" scene
3. Hit play
4. When the video is being played press "space" on the keyboard and observe a profiler
Expected: After hitting space VideoPlayer.Stop() method is called and video is stopped.
Actual: After VideoPlayer.Stop() method is called there is a performance spike, FPS drops from ~90 to ~15.
Reproduced with:5.6.0b1, 5.6.2p1, 2017.1.0f2, 2017.2.0b1
Note: Issue is reproducible only on Windows OS (tested with Windows 10 64-bit)
Fixed in: 2017.3.0a7
Backported to 2017.2.0b5
Comments (4)
Add comment
All about bugs
View bugs we have successfully reproduced, and vote for the bugs you want to see fixed most urgently.
Latest issues
- Keyboard navigation not working well with screen readers on mobile platforms
- Crash on SpriteAtlas::PopulatePackedTexturesAndRenderData when changing Import Settings on a Sprite that has been added to a Sprite Atlas while in PlayMode
- Empty lists can be applied in Additional Compiler Argument list in Project Settings
- [Vulkan][OpenXR] SRP Foveation Regions Not Centered when eye-tracking is not used
- Long Color Library name overflows outside input field bounds
AlexisDiStefano
Aug 02, 2017 19:47
:/ same here
JDMulti
Jul 31, 2017 08:46
I've reported the same issue and told they're looking into it. At the moment using the videoplayer for VR is a no-go unless this is fixed. Stopping a video and see a framedrop isn't acceptable for VR. Hopefully they can fix it soon as this videoplayer is a brand new feature.
I reproduced the issue in 2017.1.0.f3.
dmart331
Jul 28, 2017 20:45
We are also seeing this in 2017.1.0f3.
xavier-heimgartner
Jul 27, 2017 08:40
I'm seeing the same issue. In my case, the function takes about 420 ms and allocates 36.5 KB of garbage.