OpenGL for OpenCPN

[bdbcat]

Hi Team....

We are exploring the idea of shifting the OCPN graphics engine over to OpenGL. Why?
a. Better performance, leveraging modern PC graphics cards that are ubiquitous and powerful.
b. Future cross-platform support (e.g Android/IOS)

So, I have created a test version of OCPN using OGL. This is initially available for linux developer types at the following repo:

git://github.com/bdbcat/OpenCPN.git

Branches: master is OCPN 2.5, opengl is present work.

Here are some release notes:
1. Build process is normal cmake/build as for main trunk OCPN.
2. libglew and libglew-dev are required. Should be available in most linux repos.
3. Main area of interest is quilted raster charts.
4. Vector charts do not work, or may work poorly.
5. Lots of other on-screen graphics items are not rendered correctly.
6. OpenGL is enabled on Toolbox->Settings. Note "smooth zoom" option.

The idea here is to do the following:

1. Evaluate compatibility across various OGL platforms. Will it build? Will it run?
2. Evaluate perceived performance on quilted raster charts. Does it "feel" faster than OCPN 2.5?

So, if any linux types have the time and inclination to clone and build this repo, I'd appreciate some rough alpha testing.
If you do test, and can report results, we would be interested in knowing:

a. System config (CPU/RAM/etc)
b. Graphics card info, especially memory size.
c. FPS results of the glxgears demo program, for reference.
d. Anything else you think relevent
e. Cool screenshots...

Next step: If any Windows developers are able to port this code to Windows, I would be happy to take patches.
I do not currently have an OGL capable Windows box. Vbox is complaining......

Finally, if any true OpenGL gurus are aboard, feel free to criticize and improve the code. I'm just building this thing ab-initio in a vacuum....
There may be much better architectural ideas for raster chart management and rendering.

Many, many thanks to Sean D'Epagnier, who created the OGL framework for OCPN, and got this branch started. Architectural errors in in the present work are mine, not his...

Have fun

Thanks
Dave

[rooiedirk]

Hoi Dave,
Did a small test on my Gentoo box.
1 Compiling goes fine
2 Unfortunately I have only vector charts available (CM93 and ENC). Both chart types where working but no (or not much) speed improvement when zooming in/out.
Dirk

[bdbcat]

DIrk...

Thanks for the test.

Results as expected. There is no improvement, and probably some degradation, of vector chart performance in this OpenGL test code.

Porting vector charts to native OpenGL rendering is in the plan, but is a big undertaking. We'll start that after we get some ideas of raster chart behavior.

Thanks
Dave

[cagney]

Dave

Testing on an eeepc 901 from 98 1.6 GHz atom processor and 1gig memory.
Very low end graphics described in a review like this

In the video depiction one finds the limits of the CPU/GPU combination. The not very strong Atom processor isn't supported by a graphic card which accelerates the video decoding. Because of this, the performance is only adequate for watching SD videos. MPEG 2 videos (DVD) could only be played with a load of 20 – 40% smoothly (depending on the energy saving mode). HD videos could also only be admired in 480p as single pictures. Web videos, depending on the data rate, can also be enjoyed. A for the test played flash video on Gamestar.de (F.E.A.R. 2 trailer) jerked in the standard settings of 1500 Kbits/s. At 768 Kbits/s the video ran smoothly.

Testing
glxgears gave this result

231 frames in 5.0 seconds = 46.064 FPS
251 frames in 5.0 seconds = 50.160 FPS
254 frames in 5.0 seconds = 50.764 FPS

Opencpn didn't manage more than barely showing the outlines of the charts, but no rendering. Zooming and panning was no go


Thomas

[cagney]

There are two branches at the git address given.
To get the opengl branch use this command

 git clone -b opengl git://github.com/bdbcat/OpenCPN.git

Thomas

[sailias]

Is there a variable that can be set so that this version can be installed side by side with the existing 2.5 release, rather then overwrite the exiting install build from source? That would be very helpful for testing purposes...<g>


Regards and thanks

[sailias]

I should mention testing on linux....

/ch

[sailias]

Compiles fine on 11.10 alpha and my asus netbook specs can be found here.

Asus EEE PC 100HA Price, Specs – 10.1-inch Netbook India

Raster charts are definitely faster on this machine and routemanager works fine with waypoints present.

The one thing that I noticed is the its still a bit rough integrating the charts. Zoom is fantastic but when it has to switch to a new chart, there is a moments hesitation where the screen turns bright pink and then the new chart appears. It may be a limitation of the horse power on this machine, but its served me fine to date.

The only difference in the stated spec is that I have 2 megs of RAM not 1. Dont know enough about openGL to tweak it on Ubuntu but Im sure there are optimisations available.

HTH's

/ch

[cagney]

Originally Posted by sailias

Is there a variable that can be set so that this version can be installed side by side with the existing 2.5 release, rather then overwrite the exiting install build from source? That would be very helpful for testing purposes...<g>


Regards and thanks

Just skip "make install". Copy the binary "opencpn" from "build" to a suitable directory and run it with a command line like

$/path/to/opencpn

Thomas

[sailias]

Thanks....should have thought of that now that you point it out.

/ch

[cagney]

Dave

Testing with a bit with more power.
Impressive improvement! Specially the zooming, but also much snappier panning.

a. System: Athlon II X3 2,9 Ghz 4 Gb memory.
b. Graphics card info: ATI HD 5450 1Gb memory
c. FPS results:301 frames in 5.0 seconds = 60.024 FPS
d. Anything else... Just upgraded to kubuntu 11.10 beta1 to get my ATI card running properly.
e. Cool screenshots... Just interesting to compare how GL is rendering the pixels.

First without GL...


and with GL


ENC and CM93 charts are not working at all.

Thomas

[bdbcat]

Thomas, and others...

Thanks for the info. It is helpful.

Technical digression:

The biggest issue for raster chart performance on OpenGL relates to "texture memory space".
We split raster charts into 512x512 tiles at their native pixel resolution, and then load them into the GPU as needed for accelerated scaling up and down. When panning, we simply change the GPU viewport mapping matrix.

All this is very fast and smooth if the tiles can stay resident on the graphics card VRAM. When this VRAM gets filled up, the GPU moves the least recently used tile(s) to system memory to make space available. Subsequent reloading to the VRAM, though automatic, is thus slower.

Bottom line: We need substantial graphics memory to do rasters well. A further complication relates to screen pixel size. I find that 64MB is not enough for a 1280x1024 screen at high underzoom. I suspect 1GB (Thomas) is overkill. For the Asus EEEPC 100HA (Sailias), the intel GMA 950 has a configurable shared memory window. I suppose the standard OGL drivers for this chip are doing a decent job of allocating resources.

Thomas, do you have a spec link for your little Atom box? Wonder what GPU it uses?

Note Thomas' screen shots. This is part of the payoff of OpenGL. Highly underzoomed charts will look good, even withough significant CPU processing. The GPU is doing the work, using something called "Mipmaps". For our purposes, mipmaps can be though of as precomputed zoom-outs of each tile, just waiting to be mapped into screen memory as needed. GPU does all the calculations and averaging between fixed zoom levels.

c++ hackers will find in chcanv.cpp some commented printf statements that try to report timing values associated with texture loading and swapping. Turn these on if you want to get some idea of the texture management overhead.

Finally, I should have mentioned before, that a report of the output of

$glxinfo

would be useful for analysis of different performance levels and compatibility.

Have Fun
Dave

[sailias]

Lots of data is output....model "1000HA" Asus, Ubuntu 11.10 alpha (with updates)
--------------------------------
chris123@segrund:~$ glxinfo
name of display: :0
display: :0 screen: 0
direct rendering: Yes
server glx vendor string: SGI
server glx version string: 1.4
server glx extensions:
GLX_ARB_multisample, GLX_EXT_import_context, GLX_EXT_texture_from_pixmap,
GLX_EXT_visual_info, GLX_EXT_visual_rating, GLX_MESA_copy_sub_buffer,
GLX_OML_swap_method, GLX_SGI_make_current_read, GLX_SGI_swap_control,
GLX_SGIS_multisample, GLX_SGIX_fbconfig, GLX_SGIX_pbuffer,
GLX_SGIX_visual_select_group, GLX_INTEL_swap_event
client glx vendor string: Mesa Project and SGI
client glx version string: 1.4
client glx extensions:
GLX_ARB_get_proc_address, GLX_ARB_multisample, GLX_EXT_import_context,
GLX_EXT_visual_info, GLX_EXT_visual_rating, GLX_EXT_framebuffer_sRGB,
GLX_MESA_copy_sub_buffer, GLX_MESA_multithread_makecurrent,
GLX_MESA_swap_control, GLX_OML_swap_method, GLX_OML_sync_control,
GLX_SGI_make_current_read, GLX_SGI_swap_control, GLX_SGI_video_sync,
GLX_SGIS_multisample, GLX_SGIX_fbconfig, GLX_SGIX_pbuffer,
GLX_SGIX_visual_select_group, GLX_EXT_texture_from_pixmap,
GLX_INTEL_swap_event
GLX version: 1.4
GLX extensions:
GLX_ARB_get_proc_address, GLX_ARB_multisample, GLX_EXT_import_context,
GLX_EXT_visual_info, GLX_EXT_visual_rating, GLX_MESA_copy_sub_buffer,
GLX_MESA_multithread_makecurrent, GLX_MESA_swap_control,
GLX_OML_swap_method, GLX_OML_sync_control, GLX_SGI_make_current_read,
GLX_SGI_swap_control, GLX_SGI_video_sync, GLX_SGIS_multisample,
GLX_SGIX_fbconfig, GLX_SGIX_pbuffer, GLX_SGIX_visual_select_group,
GLX_EXT_texture_from_pixmap, GLX_INTEL_swap_event
OpenGL vendor string: Tungsten Graphics, Inc
OpenGL renderer string: Mesa DRI Intel(R) 945GME x86/MMX/SSE2
OpenGL version string: 1.4 Mesa 7.11
OpenGL extensions:
GL_ARB_multisample, GL_EXT_abgr, GL_EXT_bgra, GL_EXT_blend_color,
GL_EXT_blend_logic_op, GL_EXT_blend_minmax, GL_EXT_blend_subtract,
GL_EXT_copy_texture, GL_EXT_polygon_offset, GL_EXT_subtexture,
GL_EXT_texture_object, GL_EXT_vertex_array, GL_EXT_compiled_vertex_array,
GL_EXT_texture, GL_EXT_texture3D, GL_IBM_rasterpos_clip,
GL_ARB_point_parameters, GL_EXT_draw_range_elements, GL_EXT_packed_pixels,
GL_EXT_point_parameters, GL_EXT_rescale_normal,
GL_EXT_separate_specular_color, GL_EXT_texture_edge_clamp,
GL_SGIS_generate_mipmap, GL_SGIS_texture_border_clamp,
GL_SGIS_texture_edge_clamp, GL_SGIS_texture_lod, GL_ARB_multitexture,
GL_IBM_multimode_draw_arrays, GL_IBM_texture_mirrored_repeat,
GL_3DFX_texture_compression_FXT1, GL_ARB_texture_cube_map,
GL_ARB_texture_env_add, GL_ARB_transpose_matrix,
GL_EXT_blend_func_separate, GL_EXT_fog_coord, GL_EXT_multi_draw_arrays,
GL_EXT_secondary_color, GL_EXT_texture_env_add,
GL_EXT_texture_filter_anisotropic, GL_EXT_texture_lod_bias,
GL_INGR_blend_func_separate, GL_NV_blend_square, GL_NV_light_max_exponent,
GL_NV_texgen_reflection, GL_NV_texture_env_combine4,
GL_SUN_multi_draw_arrays, GL_ARB_texture_border_clamp,
GL_ARB_texture_compression, GL_EXT_framebuffer_object,
GL_EXT_texture_env_dot3, GL_MESA_window_pos, GL_NV_packed_depth_stencil,
GL_NV_texture_rectangle, GL_NV_vertex_program, GL_ARB_depth_texture,
GL_ARB_shadow, GL_ARB_texture_env_combine, GL_ARB_texture_env_crossbar,
GL_ARB_texture_env_dot3, GL_ARB_texture_mirrored_repeat,
GL_ARB_window_pos, GL_EXT_stencil_two_side, GL_EXT_texture_cube_map,
GL_NV_vertex_program1_1, GL_APPLE_client_storage, GL_APPLE_packed_pixels,
GL_APPLE_vertex_array_object, GL_ARB_draw_buffers,
GL_ARB_fragment_program, GL_ARB_fragment_shader, GL_ARB_shader_objects,
GL_ARB_vertex_program, GL_ARB_vertex_shader, GL_ATI_draw_buffers,
GL_ATI_texture_env_combine3, GL_EXT_shadow_funcs, GL_EXT_stencil_wrap,
GL_MESA_pack_invert, GL_MESA_ycbcr_texture, GL_ARB_half_float_pixel,
GL_ARB_point_sprite, GL_ARB_shading_language_100, GL_ARB_sync,
GL_ARB_texture_non_power_of_two, GL_ARB_vertex_buffer_object,
GL_ATI_blend_equation_separate, GL_EXT_blend_equation_separate,
GL_OES_read_format, GL_ARB_pixel_buffer_object, GL_ARB_texture_rectangle,
GL_EXT_pixel_buffer_object, GL_EXT_texture_rectangle,
GL_ARB_framebuffer_object, GL_EXT_framebuffer_blit,
GL_EXT_framebuffer_multisample, GL_EXT_packed_depth_stencil,
GL_APPLE_object_purgeable, GL_ARB_vertex_array_object,
GL_ATI_separate_stencil, GL_EXT_gpu_program_parameters,
GL_EXT_texture_env_combine, GL_OES_EGL_image, GL_ARB_copy_buffer,
GL_ARB_map_buffer_range, GL_EXT_separate_shader_objects,
GL_ARB_ES2_compatibility, GL_ARB_draw_elements_base_vertex,
GL_ARB_explicit_attrib_location, GL_ARB_provoking_vertex,
GL_ARB_sampler_objects, GL_EXT_provoking_vertex, GL_ARB_robustness

32 GLX Visuals
visual x bf lv rg d st colorbuffer sr ax dp st accumbuffer ms cav
id dep cl sp sz l ci b ro r g b a F gb bf th cl r g b a ns b eat
----------------------------------------------------------------------------
0x021 24 tc 0 32 0 r y . 8 8 8 8 . . 0 24 8 0 0 0 0 0 0 None
0x022 24 dc 0 32 0 r y . 8 8 8 8 . . 0 24 8 0 0 0 0 0 0 None
0x08d 24 tc 0 24 0 r . . 8 8 8 0 . . 0 0 0 0 0 0 0 0 0 None
0x08e 24 tc 0 24 0 r y . 8 8 8 0 . . 0 0 0 0 0 0 0 0 0 None
0x08f 24 tc 0 24 0 r y . 8 8 8 0 . . 0 0 0 0 0 0 0 0 0 None
0x090 24 tc 0 24 0 r . . 8 8 8 0 . . 0 24 8 0 0 0 0 0 0 None
0x091 24 tc 0 24 0 r y . 8 8 8 0 . . 0 24 8 0 0 0 0 0 0 None
0x092 24 tc 0 24 0 r y . 8 8 8 0 . . 0 24 8 0 0 0 0 0 0 None
0x093 24 tc 0 32 0 r . . 8 8 8 8 . . 0 0 0 0 0 0 0 0 0 None
0x094 24 tc 0 32 0 r y . 8 8 8 8 . . 0 0 0 0 0 0 0 0 0 None
0x095 24 tc 0 32 0 r y . 8 8 8 8 . . 0 0 0 0 0 0 0 0 0 None
0x096 24 tc 0 32 0 r . . 8 8 8 8 . . 0 24 8 0 0 0 0 0 0 None
0x097 24 tc 0 24 0 r y . 8 8 8 0 . . 0 24 8 0 0 0 0 0 0 None
0x098 24 tc 0 24 0 r y . 8 8 8 0 . . 0 24 8 16 16 16 0 0 0 Slow
0x099 24 tc 0 32 0 r y . 8 8 8 8 . . 0 24 8 0 0 0 0 0 0 None
0x09a 24 tc 0 32 0 r y . 8 8 8 8 . . 0 24 8 16 16 16 16 0 0 Slow
0x09b 24 dc 0 24 0 r . . 8 8 8 0 . . 0 0 0 0 0 0 0 0 0 None
0x09c 24 dc 0 24 0 r y . 8 8 8 0 . . 0 0 0 0 0 0 0 0 0 None
0x09d 24 dc 0 24 0 r y . 8 8 8 0 . . 0 0 0 0 0 0 0 0 0 None
0x09e 24 dc 0 24 0 r . . 8 8 8 0 . . 0 24 8 0 0 0 0 0 0 None
0x09f 24 dc 0 24 0 r y . 8 8 8 0 . . 0 24 8 0 0 0 0 0 0 None
0x0a0 24 dc 0 24 0 r y . 8 8 8 0 . . 0 24 8 0 0 0 0 0 0 None
0x0a1 24 dc 0 32 0 r . . 8 8 8 8 . . 0 0 0 0 0 0 0 0 0 None
0x0a2 24 dc 0 32 0 r y . 8 8 8 8 . . 0 0 0 0 0 0 0 0 0 None
0x0a3 24 dc 0 32 0 r y . 8 8 8 8 . . 0 0 0 0 0 0 0 0 0 None
0x0a4 24 dc 0 32 0 r . . 8 8 8 8 . . 0 24 8 0 0 0 0 0 0 None
0x0a5 24 dc 0 32 0 r y . 8 8 8 8 . . 0 24 8 0 0 0 0 0 0 None
0x0a6 24 dc 0 24 0 r y . 8 8 8 0 . . 0 24 8 0 0 0 0 0 0 None
0x0a7 24 dc 0 24 0 r y . 8 8 8 0 . . 0 24 8 16 16 16 0 0 0 Slow
0x0a8 24 dc 0 32 0 r y . 8 8 8 8 . . 0 24 8 0 0 0 0 0 0 None
0x0a9 24 dc 0 32 0 r y . 8 8 8 8 . . 0 24 8 16 16 16 16 0 0 Slow
0x05c 32 tc 0 32 0 r y . 8 8 8 8 . . 0 24 8 0 0 0 0 0 0 None

48 GLXFBConfigs:
visual x bf lv rg d st colorbuffer sr ax dp st accumbuffer ms cav
id dep cl sp sz l ci b ro r g b a F gb bf th cl r g b a ns b eat
----------------------------------------------------------------------------
0x05d 0 tc 0 16 0 r . . 5 6 5 0 . . 0 0 0 0 0 0 0 0 0 None
0x05e 0 tc 0 16 0 r y . 5 6 5 0 . . 0 0 0 0 0 0 0 0 0 None
0x05f 0 tc 0 16 0 r y . 5 6 5 0 . . 0 0 0 0 0 0 0 0 0 None
0x060 0 tc 0 16 0 r . . 5 6 5 0 . . 0 16 0 0 0 0 0 0 0 None
0x061 0 tc 0 16 0 r y . 5 6 5 0 . . 0 16 0 0 0 0 0 0 0 None
0x062 0 tc 0 16 0 r y . 5 6 5 0 . . 0 16 0 0 0 0 0 0 0 None
0x063 24 tc 0 24 0 r . . 8 8 8 0 . . 0 0 0 0 0 0 0 0 0 None
0x064 24 tc 0 24 0 r y . 8 8 8 0 . . 0 0 0 0 0 0 0 0 0 None
0x065 24 tc 0 24 0 r y . 8 8 8 0 . . 0 0 0 0 0 0 0 0 0 None
0x066 24 tc 0 24 0 r . . 8 8 8 0 . . 0 24 8 0 0 0 0 0 0 None
0x067 24 tc 0 24 0 r y . 8 8 8 0 . . 0 24 8 0 0 0 0 0 0 None
0x068 24 tc 0 24 0 r y . 8 8 8 0 . . 0 24 8 0 0 0 0 0 0 None
0x069 24 tc 0 32 0 r . . 8 8 8 8 . . 0 0 0 0 0 0 0 0 0 None
0x06a 24 tc 0 32 0 r y . 8 8 8 8 . . 0 0 0 0 0 0 0 0 0 None
0x06b 24 tc 0 32 0 r y . 8 8 8 8 . . 0 0 0 0 0 0 0 0 0 None
0x06c 24 tc 0 32 0 r . . 8 8 8 8 . . 0 24 8 0 0 0 0 0 0 None
0x06d 24 tc 0 32 0 r y . 8 8 8 8 . . 0 24 8 0 0 0 0 0 0 None
0x06e 32 tc 0 32 0 r y . 8 8 8 8 . . 0 24 8 0 0 0 0 0 0 None
0x06f 0 tc 0 16 0 r y . 5 6 5 0 . . 0 16 0 0 0 0 0 0 0 None
0x070 0 tc 0 16 0 r y . 5 6 5 0 . . 0 16 0 16 16 16 0 0 0 Slow
0x071 24 tc 0 24 0 r y . 8 8 8 0 . . 0 24 8 0 0 0 0 0 0 None
0x072 24 tc 0 24 0 r y . 8 8 8 0 . . 0 24 8 16 16 16 0 0 0 Slow
0x073 24 tc 0 32 0 r y . 8 8 8 8 . . 0 24 8 0 0 0 0 0 0 None
0x074 24 tc 0 32 0 r y . 8 8 8 8 . . 0 24 8 16 16 16 16 0 0 Slow
0x075 0 dc 0 16 0 r . . 5 6 5 0 . . 0 0 0 0 0 0 0 0 0 None
0x076 0 dc 0 16 0 r y . 5 6 5 0 . . 0 0 0 0 0 0 0 0 0 None
0x077 0 dc 0 16 0 r y . 5 6 5 0 . . 0 0 0 0 0 0 0 0 0 None
0x078 0 dc 0 16 0 r . . 5 6 5 0 . . 0 16 0 0 0 0 0 0 0 None
0x079 0 dc 0 16 0 r y . 5 6 5 0 . . 0 16 0 0 0 0 0 0 0 None
0x07a 0 dc 0 16 0 r y . 5 6 5 0 . . 0 16 0 0 0 0 0 0 0 None
0x07b 24 dc 0 24 0 r . . 8 8 8 0 . . 0 0 0 0 0 0 0 0 0 None
0x07c 24 dc 0 24 0 r y . 8 8 8 0 . . 0 0 0 0 0 0 0 0 0 None
0x07d 24 dc 0 24 0 r y . 8 8 8 0 . . 0 0 0 0 0 0 0 0 0 None
0x07e 24 dc 0 24 0 r . . 8 8 8 0 . . 0 24 8 0 0 0 0 0 0 None
0x07f 24 dc 0 24 0 r y . 8 8 8 0 . . 0 24 8 0 0 0 0 0 0 None
0x080 24 dc 0 24 0 r y . 8 8 8 0 . . 0 24 8 0 0 0 0 0 0 None
0x081 24 dc 0 32 0 r . . 8 8 8 8 . . 0 0 0 0 0 0 0 0 0 None
0x082 24 dc 0 32 0 r y . 8 8 8 8 . . 0 0 0 0 0 0 0 0 0 None
0x083 24 dc 0 32 0 r y . 8 8 8 8 . . 0 0 0 0 0 0 0 0 0 None
0x084 24 dc 0 32 0 r . . 8 8 8 8 . . 0 24 8 0 0 0 0 0 0 None
0x085 24 dc 0 32 0 r y . 8 8 8 8 . . 0 24 8 0 0 0 0 0 0 None
0x086 24 dc 0 32 0 r y . 8 8 8 8 . . 0 24 8 0 0 0 0 0 0 None
0x087 0 dc 0 16 0 r y . 5 6 5 0 . . 0 16 0 0 0 0 0 0 0 None
0x088 0 dc 0 16 0 r y . 5 6 5 0 . . 0 16 0 16 16 16 0 0 0 Slow
0x089 24 dc 0 24 0 r y . 8 8 8 0 . . 0 24 8 0 0 0 0 0 0 None
0x08a 24 dc 0 24 0 r y . 8 8 8 0 . . 0 24 8 16 16 16 0 0 0 Slow
0x08b 24 dc 0 32 0 r y . 8 8 8 8 . . 0 24 8 0 0 0 0 0 0 None
0x08c 24 dc 0 32 0 r y . 8 8 8 8 . . 0 24 8 16 16 16 16 0 0 Slow
----------------------------------------------------eof---------------------

Regards

/ch

[sailias]

One thing I noticed in the opengpl version and not in 2.5 is that when a map is not present, the collars are presented. This is not the case (I believe) in 2.5 or at least not seen before.

/ch

[bdbcat]

Thomas....

I'd like to investigate your eeepc901. Why does it not work?

I read that it uses the intel GMA950 graphics chip. Maybe it is a memory issue....

Could you post results of glxinfo, and also contents of /etc/X11/xorg.conf?
Anything in the Xorg log that looks interesting....

Thanks
Dave