Android development for humanity.
You can contact me via: omergilad@gmail.com
|
May
21 |
|
comment |
Optimizing performance of a heavy fragment shader It will be the same algorithm. The gather per output pixel does the offsetting according to the original texture pixel size, and the filtering is GL_NEAREST. So effectively, a bunch of different fragments fall on the same original texels. Some of the calculations will still be relevant to each fragment though and the'll be done in the fragment shader. |
|
May
21 |
|
awarded | Enthusiast |
|
May
20 |
|
comment |
Optimizing performance of a heavy fragment shader I fail to see what am I supposed to do with this. Is that considered "documentation"? |
|
May
20 |
|
comment |
Optimizing performance of a heavy fragment shader @ahcox - What about the fact that: 1. My CPU is at 0% during this execution and the GPU is a big bottleneck, and 2. Everything this fragment shader does will be done 4x-16x times less than now (depending on scaling factor). |
|
May
20 |
|
comment |
Optimizing performance of a heavy fragment shader This link gives the general idea: board.byuu.org/viewtopic.php?f=10&t=2248, but it's not the exact implementation I use here. |
|
May
20 |
|
comment |
Optimizing performance of a heavy fragment shader I did not find any documentation via Google. Can you link? |
|
May
20 |
|
answered | android: how to read json data |
|
May
20 |
|
comment |
Optimizing performance of a heavy fragment shader No OpenGL ES 3 for me... I support 2. |
|
May
20 |
|
comment |
Custom Camera App crashes in some devices occasionally Having a log or stack trace would be helpful. Ask whoever has the device to email you a log right after the crash using a third party app. |
|
May
19 |
|
comment |
Optimizing performance of a heavy fragment shader Googled it a bit - isn't the extension you mentioned just a way to use an FBO in older API versions? |
|
May
19 |
|
comment |
Optimizing performance of a heavy fragment shader Thanks, I'll look into that. Is the extension you mentioned applicable to most GPUs, or just Mali series? |
|
May
19 |
|
comment |
Optimizing performance of a heavy fragment shader How do you suggest to apply what you said? I checked for max varyings on that GPU, used all of them (as vec2) for passing texture coordinates and sampled with them, and sampled the rest of the texels with the normal offset calculations. That didn't help, but maybe I did something wrong. |
|
May
19 |
|
comment |
Optimizing performance of a heavy fragment shader I tried using multiple vec2 varying for the lookups (as much varyings as possible) and do the rest of the lookups the normal way. It just got slower... |
|
May
19 |
|
comment |
What are some ways to prevent or reduce cheating in online multiplayer games? Also, if you patch your game frequently and change binary offsets in the executable or insert random NOPs, no hack will survive long enough... |
|
May
18 |
|
awarded | Editor |
|
May
18 |
|
revised |
Optimizing performance of a heavy fragment shader added 189 characters in body |
|
May
18 |
|
comment |
Optimizing performance of a heavy fragment shader It doesn't seem separable - but I'm open to any other 2-pass suggestions that I didn't think of yet. If I had a way to output multiple intermediate values (per pixel) to one texture and then use them in another pass that would be great. |
|
May
18 |
|
comment |
Android - How to track app Resumes only when you exit and come back to the app? This is the correct way... Your "events" take place when activities_num == 0 (background) or activities_num == 1 (foreground). |
|
May
18 |
|
awarded | Student |
|
May
18 |
|
comment |
Optimizing performance of a heavy fragment shader I thought of an unorthodox way of eliminating my texture2D calls - draw a grid of vertices instead of just one quad, and pass to each one of them attributes containing all the needed texels. That would eliminate all texture sampling, but create a large drawing buffer every frame. Before I attempt to implement this lengthy solution, any advice? |
