Fix for 2.66a
""We can't pass the results of canvas->GetViewPort() directly because canvas->SetViewPort() does some extra math""
Bug introduced during 2.65 series in the refactor to use canvas->SetViewPort instead of direct opengl calls for viewport
(53305, 53392, 53393)
Fix for 2.66a
With help from Daniel Stokes and Mitchell Stokes.
This bug always existed in OSX, but started showing up in Windows and Linux on review (54745 + 54747)
[the patch to enable alpha buffer for all OSs]
A better fix would be to use RAS_IRasterizer::SetAlphaBlend(GPU_BLEND_SOLID);
but I think gpu_verify_alpha_blend() is not switching to SOLID because
GTS.alphablend is GPU_BLEND_SOLID (even though GL_ALPHA_TEST is enabled).
Anyways, this is not something worth tackling now, since in terms of functionality it shouldn't matter.
Instead of flagging the rigid body world for frame update just call
BKE_rigidbody_do_simulation() recursively for all scenes.
This avoids having to constantly check if the simulation needs to be
updated.
The rigid body world could be rebuilt on start frame and one frame after
start frame. The latter was necessary sice animation playback usually
doesn't start at start frame.
This lead to different simulations depending on which frame the
simulaton was rebuilt when animation was involved.
Now we only rebuild the world on start frame.
This is actually tricky to do since, as mentioned above, animation
playback starts on second frame. To work around this we rebuild the
world before the actual update.
The alternative would be to rebuld the world on every simulation change
(like the other simulations do it) but this is an expensive operation
and would be too slow.
to support more than 2 UV maps. This code indirectly depended on the order of
OpenGL attribute ID's assigned by the OpenGL driver being the same as the
attributes being declared in the GLSL shader code, which is not always the case.
Fix for 2.66a
Screencast could be started twice in a row, but it then also added two handlers
for drawing an overlay circle around mouse cursor. After ending the screencast,
this circle then kept being drawn.
Added compatibility option "Straight Alpha Output" to image input node
When this option is enabled, image input node will convert float buffer
to straight alpha.
This is not what you'll usually want with new alpha pipeline, nit this
is needed to preserve compatibility with older files saved in 2.65.
In that version byte image are resulting with straight alpha passing
to the compositor and alpha-overing required extra premultiplication
of inputs.
So, that's why Straight Alpha Output is needed -- it's set in versioning
code for byte node images so they'll still output straight alpha.
This option is currently only available in N-panel.
Additional change: added Alpha Mode for image input node to N-panel.
- optional, select between name/keybinding.
- when key input can't be parsed, alert red to give the user some feedback.
- key shortcut parsing could be improved or swapped out for button which grabs shortcut.
and use_expand properly. These are hardcoded to shift and ctrl to make them
work when clicking buttons or menus. Now it checks if the properties are set,
which is still not ideal but makes it possible to override them from the
key configuration.
Now the RenderEngine.render callback allows writing blender data again, it
should not be allowed but in practice the API and render threading code is
too limited to make this work at the moment.
Before alpha channel was simply ignored causing bad looking
straight colors which is pretty much useless.
Now saving RGB would alpha-over image on top of black color,
which makes final image look really nice. It's also very
such the same what other graphics software does this.
In the future we could easily support configurable backdrop
color, which would be really the same as other SW does it.
Also, it'll probably worth adding the same mode to RGB
display of image editor.
This makes it easier and useful to investigate alpha channels,
without this straight colors are displayed on the screen which
are completely useless in case of player.
Issue happened for scene. movie clip and mask strips, which contains
pointers to datablocks which are freeing on loading new file.
Also, scene strip would crash when pasted from clipboard after scene
was unlinked from file.
Such kind of image sequences wouldn't be displayed still
and supporting them would take some additional time, but
for now it'll be no memory leak on attempt opening such
images.
Several major things are done in this commit:
- First of all, logic of modal solver was changed.
We do not rely on only minimizer to take care of
guessing rotation for frame, but we're using
analytical rotation computation for point clouds
to obtain initial rotation.
Then this rotation is being refined using Ceres
minimizer and now instead of minimizing average
distance between points of point of two clouds,
minimization of reprojection error of point
cloud onto frame happens.
This gives quite a bit of precision improvement.
- Second bigger improvement here is using bundle
adjustment for a result of first step when we're
only estimating rotation between neighbor images
and reprojecting markers.
This averages error across the image sequence
avoiding error accumulation. Also, this will
tweak bundles themselves a bit for better match.
- And last bigger improvement here is support of
camera intrinsics refirenment.
This allowed to significantly improve solution
for real-life footage and results after such
refining are much more usable than it were before.
Thanks to Keir for the help and code review.
