http://www.mediafire.com/file/ekh3vsifyxutsaj/
Past Builds:
Release
https://www.mediafire.com/folder/jqqg6j ... l+Releases
Debug
https://www.mediafire.com/folder/imf9hw ... odel+Debug
Feel free to let Bart and Ian know about the bugs and glitches happening in any builds in that topic!
SVN commit info:
https://sourceforge.net/p/model3emu/code/commit_browser
- Code: Select all
r886
values must be signed, to handle negative numbers, my bad
---------------------
r885
minor fix
---------------------
r884
Added configurable refresh rate. Default is 60 but -true-hz on the command line enables 57.524 Hz mode, which some players report making Spikeout and the timing of combos in VF3 more authentic. The config key is RefreshRate and is specified as a floating point value. Should the desire arise, arbitrary refresh rates can be set in Supermodel.ini using this key but almost no validation of the value is performed. Thanks to forum user trap15 for the original code and PonMi for advocating for this patch to be included.
---------------------
r883
njz3: Proper quadrophonic audio support. Need 4+ speakers to take advantage otherwise down mixed to stereo.
---------------------
r882
Added default values for PortIn, PortOut, and AddressOut
---------------------
r881
fix: explicit specialization requires 'template <>'
---------------------
r880
Bitmap size field in the BITMAPV4HEADER was not being set correctly (the image viewers I've been using are unaffected by this but it ought to be corrected)
---------------------
r879
When dumping textures, two T1RGB5 texture maps are written: with and without contour processing. Contour processing can be disabled per-polygon and some textures intended to be used without contour processing contain pixels with T=1 that are clearly not supposed to be transparent. We now output textures_t1rgb5_contour.bmp and textures_t1rgb5_opaque.bmp.
---------------------
r878
Added a -dump-textures option (config key DumpTextures) that writes texture BMP files, one for each known format (12 in all currently)
---------------------
r877
Adjust the number of PowerPC cycles executed per frame to assume 57.524 Hz refresh rate
---------------------
r876
Double buffer the line of sight values. Should allow the correct values to be returned if GPU threading is enabled.
---------------------
r875
Fixing Magical Truck Adventure bad ROM dump which was causing attract mode to stop rendering
---------------------
r874
methods should be marked as noexcept because they cannot throw
---------------------
r873
white space.
---------------------
r872
Better document real3d registers
---------------------
r871
Clear all Real3D DMA registers on reset; uninitialized values were causing issues with MSVC debug builds
---------------------
r870
Another header include fix
---------------------
r869
Include required header files in Model3GraphicsState.h - fixes potential build errors in GCC
---------------------
r868
Increase MIDI FIFO buffer size for SCSP; should prevent FIFO overflows
---------------------
r867
PowerPC IRQ line deassertion; DMA only fires interrupts when required
---------------------
r866
Getting rid of most of the includes from Supermodel.h; each file now explicitly includes the header files it needs.
---------------------
r865
Shared RAM on simulated netboard is now double-buffered; Spikeout Final Edition no longer requires a hack to work.
---------------------
r864
New JTAG patch for Sega Rally 2; linked mode now fully working.
---------------------
r863
Increase size of DSB2 FIFO buffer; fixes Sega Rally 2 music occasionally not switching or fading out
---------------------
r862
Non-linked games no longer falsely list the net board under extra hardware
---------------------
r861
Fix 16-bit variables not displaying correctly in Z80 debugger (at least in MSVC builds)
---------------------
r860
Fix netboard resets and detection of simulated netboard linking to itself
---------------------
r859
Adding simulated netboard; all linked games except Sega Rally 2 working at full speed.
---------------------
r858
Corrected alignment issues with netboard RAM; send/receive parameters no longer need to be manipulated to work correctly. Fixes bugs including Scud Race linked car orientation.
---------------------
r857
Add some async network methods. Async send copies to an internal buffer and sends in another thread so doesn't block at all. Added CheckDataAvailable method to the TCPReceive class. Can check or wait for data without blocking.
---------------------
r856
DSB2 clock speed from 4MHz to 11MHz
---------------------
r855
New dumps added:
Virtua Striker 2 '99.1 (Step 1.5, Export, USA, Revision B)
Virtua Striker 2 '99.1 (Step 1.5, Japan, Revision B)
Thanks to Brian Troha and The Dumping Union
---------------------
r854
Fixed cycle counting overflow that was causing DSB2 to halt execution.
---------------------
r853
missing header
---------------------
r852
DSB2 IRQ 2 now fires at 1KHz rather than once per frame, improving
music timing in Daytona USA 2 and Sega Rally 2. DSB1 CPU timing
increased from 1MHz to 4MHz, improving music fade timing in Scud Race.
Thanks to gm_matthew for these discoveries.
---------------------
r851
Just a few compile warnings
---------------------
r850
Fix project file
---------------------
r849
Build bot script.
---------------------
r848
Skichamp Driveboard error fix.
---------------------
r847
Huge refactor of the Driveboard:
---------------------
r846
Introduce new config keywords.
---------------------
r845
The controls of bass/getbass games were no longer good, following recent rom set name changes in Games.xml.
---------------------
r844
Corrected game set names & game titles (Brian Troha)
---------------------
r843
Added rumble skipad in Ski Champ game.
---------------------
r842
Service menu activated in swtrilgyp.
---------------------
r841
- Add -O3 optimization option for OSX and Linux.
---------------------
r840
- Suppressed net board console output (using debug logging now)
---------------------
r839
- Star Wars Trilogy untextured Death Star surface bug in gcc builds was caused by -Ofast (thanks to rokfpoewrkcpoqwkcp for discovering this). Using -O3 now.
---------------------
r838
Updated rules file.
---------------------
r837
Introducing a special release build target with SVN version stamping.
---------------------
r836
Magical Truck Adventure rom patch to unlock region. (enter service menu, then Start P1, Start P1, Service, Start P1, Service, Test)
---------------------
r835
-wide-bg option to stretch the background tile layer when wide-screen mode is enabled - Daro Land
---------------------
r834
Change the duration of constant force effect in SDL2 to correct daytona games
---------------------
r833
Makefiles update
---------------------
r832
Star Wars Trilogy lever feedback activated
---------------------
r831
Force Feeback for Linux
---------------------
r830
hopefully fix building on linux
---------------------
r829
Added Screenshot feature (ALT+S).
---------------------
r828
Remove using namespace from headers.
---------------------
r827
fix compiling with visual studio
---------------------
r826
fix missing polys with quad renderer
---------------------
r825
Balance initial value set to "0" rather than "false". "False" is interpreted as 0 when decoded as a float, so there will be no effect here, but this expresses the intended logic better.
---------------------
r824
Fixed a longstanding bug that caused stereo channels to be flipped incorrectly.
---------------------
r823
Clip scroll fog to viewport, instead of filling entire screen. Fixes fog bug in Daytona credits.
---------------------
r822
Update scroll fog logic. Should fix the credits in virtua fighter and daytona that have fogging applied to the background layer and shouldn't.
---------------------
r821
- Add missing library from previous windows makefile update.
---------------------
r820
Windows makefile update
---------------------
r819
- Reworked logging system to support output to multiple outputs: files, stdout, stderr, and syslog (OutputDebugString on Windows, syslog on other systems).
---------------------
r818
Rename command line SCSP option to -legacy-scsp and -new-scsp
---------------------
r817
Correct the returned value from previous getbass patch
---------------------
r816
Getbass related
---------------------
r815
Small oversight in the new SCSP code
---------------------
r814
Various SCSP improvements and code cleanup
---------------------
r813
New ecap crc rom + rom patch
---------------------
r812
Low level z80 core bug
---------------------
r811
Remove some debug code
---------------------
r810
Update glew version
---------------------
r809
build fixes for mac
---------------------
r808
skichamp rom patch : prevent driveboard error at boot
---------------------
r807
Linux build compatibility changes
---------------------
r806
Update makefiles
---------------------
r805
silence a few casting warnings
---------------------
r804
Netboard related : Define a new hardware line in games.xml
---------------------
r803
_rotl() replacement
---------------------
r802
Only try and make a connection if we have enabled EmulateNet
---------------------
r801
Fix corrupted memory in network mode
---------------------
r800
remove old files
---------------------
r799
Replace network code with new tcp implementation
---------------------
r798
Add SDLnetworking + some basic tcp classes
---------------------
r797
cache variable, without it is unusably slow in debug mode
---------------------
r796
add missing file
---------------------
r795
update vs project files
---------------------
r794
fix output values
---------------------
r793
The translator map seems to work with paletted colour values too, and the result is not clamped. Colours are passed to the GPU as unsigned bytes to multiplying by 16 will overflow, so we do the logic in the shader. If we passed floats we could skip the shader logic.
---------------------
r792
When the translator map is enabled the colour values seem to use 4 bits, 0-16. Why not 0-31 I don't know. Anyway virtua on is overflowing these values for some of the smoke effects. For us the numbers were wrapping around. Clamping is enough to fix this. (Thanks model123)
---------------------
r791
custom resolution bug fix
---------------------
r790
SDL1 to SDL2 internal changes
---------------------
r789
remove files
---------------------
r788
---------------------
r787
tweak value so the map indicator still works at the start of spikeout
---------------------
r786
Sutherland-Hodgeman clipping only works for planar primitives. Scud is rendering non planar quads and this breaks our algorithm a little when using the quad renderer. Stretching the near plane a few % is enough to fix it.
---------------------
r785
Only need to clip against 4 planes for poly culling
---------------------
r784
Fix the sky in harley in first person mode where pixels in the sky are culled when they shouldn't be with the quad renderer. Basically we were running out of precision in a rare corner case. Doing the maths with double precision and casting back to float was enough to fix the issue.
---------------------
r783
To calculate the mipmap value the opengl needs to calculate the difference in the texture coordinates between adjacent pixels. If there are no adjacent pixels the hw may need to spawn invisible fragments outside of the visible polygon. Our algorithm was culling these out so the mipamp calculation failed for edge pixels. Anyway this fixes quad rendering looking broken with the latest nvidia drivers.
---------------------
r782
Really should cull against 5 frustum planes.
---------------------
r781
Stretch the time the ping_pong bit flips a bit to bump writes into correct frame. Otherwise they end up getting written at like 99.8% of the frame and overlapping with the next.
---------------------
r780
Optimise Games.xml (Krom)
---------------------
r779
Thanks to our anonymous contributor who managed to patch a model 3 game and run it on real h/w, we were able to obtain some timing values we had been long been missing for correct emulation. The h/w polls something called the ping_pong bit at start-up to sync the GPU with the CPU. Unfortunately we didn't have the correct values and just manually used some per game hacks to get games to run. This mostly worked, but some games were writing more or less frames than they should have been for a given time period. When the ping_pong bit flips at 66% of the frame time, games were writing data for a new frame, which meant writes were often straddling 2 separate frames. We aren't 100% sure if IRQ2 or the ping_pong bit is vblank.
---------------------
r778
Ocean hunter in the middle of the game is passing a few matrices with FLT_MAX inside them, which blows apart our near/far calculation. The hardware must have some method to sanitize the near/far for instance if you render something extremely close to the origin you will also blow apart any near/far calc.
---------------------
r777
skip invalid culling nodes
---------------------
r776
A culling node must follow immediately after a viewport.
---------------------
r775
Fix possible negative pointer arithmetic which was causing the music to wrap around.
---------------------
r774
Update makefiles
---------------------
r773
Previous mpeg audio code worked for me in debug mode, but the release build had some serious corruption issues when being built in visual studio. Replaced the old spaghetti code with a more modern implementation.
---------------------
r772
- Disabled Alt-O (dump timings) and Alt-U (dump input state) when
---------------------
r771
Restore Sega Rally 2 music in some practice stages
---------------------
r770
Fixed MPEG music looping glitch.
---------------------
r769
Refactor code from previous svn768 (Note : break old savestate compatibility)
---------------------
r768
Fix the audio in sega rally2. The game has packed separate tracks into the left and right audio channels. Selecting the correct track fixes the garbled mess. (Spindizzi)
---------------------
r767
Fix shader. Why this worked on my nvidia card .. I'm not sure.
---------------------
r766
Composite the alpha layers at the end of rendering. To do this we need to mask the alpha pixels with the opaque pixels from the next priority layer. Fixes some overlapping shadows in vf3tb that have different priority layers. I assume that was a game bug, but it works on the real h/w.
---------------------
r765
-Added support for specifying multiple sections simultaneously in INI
---------------------
r764
Technically backface culling should cull polys when the dot product = 0, since when it equals 0 the polys is perpendicular to the camera and thus invisible. When you complete sega rally2 (if you can get that far lol) the champagne is invisible. The reason is because all the face normals are 0. So when multiplied by the model matrix they are still 0 and get culled. Tweaking the condition to only cull polys when greater than 0 fixes this, and allows these 'bad' polys to render as they do on the model3.
---------------------
r763
Inputs can now be configured on a per-game basis
---------------------
r762
- PCI bridge can be specified in Games.xml for games that use a different version than expected based on stepping. No more hard-coded exceptions.
---------------------
r761
timing fix for srally2x (Spindizzi)
---------------------
r760
Updates games.xml for new prototype games (Spindizzi)
---------------------
r759
remove commands that do nothing
---------------------
r758
clear scissored out areas
---------------------
r757
Optimise depth calculation for quad shader.
---------------------
r756
Update Games.xml for games that were recently dumped
---------------------
r755
---------------------
r754
Fade out microtextures with higher textures LODs. Fixes incorrectly applied microtextures in LA machineguns.
---------------------
r753
debug code should be disabled
---------------------
r752
Smooth texture repeat makes no sense for alpha/contour textures with pixel dilate. HW seems to treat them as non smooth anyway. Also implement some line of sight stuff, used by Scud. Really need to check the threading/synchronisation of the register reads but it basically works.
---------------------
r751
return correct PCI ID
---------------------
r750
fix debugging code
---------------------
r749
fix copy paste error :]
---------------------
r748
Correctly emulate the real3d pro-1000 texture modes in our shader.
---------------------
r747
Revert some patches (Spindizzi)
---------------------
r746
Update makefile
---------------------
r745
Both lemans24 and dirt devils are rendering some totally transparent objects as the first model in each viewport. The HW seems to skip these entirely, and including them breaks our near/far plane calculations leading to Z fighting.
---------------------
r744
Tweak the ratio for the near plane. Fixes missing text on the map indicator at the start of spikeout. The vertices are 0.003 units apart so a tonne of precision is required at the near plane to not depth test out.
---------------------
r743
remove debug code
---------------------
r742
avoid double matrix load
---------------------
r741
updated command line help
---------------------
r740
Quad rendering engine. Set QuadRendering = 1 in the ini file to use, or -quad-rendering at the end of the command line to use.
---------------------
r739
Sega bass fishing is starting some off the meshes with a shared vertex. Without having a previous vertex to share with, the values were just uninitialised and junk. The hardware itself seems to hold the value from the previous mesh, and sega bass fishing relies on this undefined behaviour to function correctly. Storing the cached vertices as member variables instead of locally to the function means it holds the values from the last mesh, and uses those values instead of random garbage.
---------------------
r738
remove dead variable
---------------------
r737
Use correct PCI IDs (Harry Tuttle)
---------------------
r736
add crosshair effect (Harry Tuttle)
---------------------
r735
fix cast, and remove extra ; that had snuck into the shaders
---------------------
r734
Opaque polys definitely depth test with less than or equal. Virtua striker is drawing co-planar polys for the bar gfx
---------------------
r733
Draw transparent polys to separate layers and composite at the end. This solves a tonne of transparency errors we had been battling with for a long time. The model3 is strange in the fact it only supports a max of two translucent overlapped polys. They are not blended into the frame normally. Doing this means the topmost translucent polys only are visible in the scene, the equivalent of doing a depth pass first, but without the added cost.
---------------------
r732
forgot parentheses
---------------------
r731
---------------------
r730
optimise hash function and parse out transparency layer select bit
---------------------
r729
select correct bit ..
---------------------
r728
Parse out TranslucencyPatternSelect
---------------------
r727
Fixed Makefiles for non-MSYS builds. Unfortunately had to revert a lot of the automated configuration that Harry had inserted. Users will again need to
---------------------
r726
Parse siblings after children. It's the only way to get Daytona water fall to look correct without h/w poly sorting
---------------------
r725
Added a stretch mode (-stretch) and modified new renderer's viewport code (hopefully correctly) to handle aspect ratio correction and widening the way the legacy one does.
---------------------
r724
Render the opaque part of texture transparency in the first pass. Fixes numerous transparency related issues in lemans24 and some in ocean hunter. I don't really know if this is the 'correct' way of solving this because the real3d pro-1000 is a complete black box. There is still a lot we don't understand about how it handles transparency.
---------------------
r723
remove unnecessary copies
---------------------
r722
Fix controller inputs (Spindizzi)
---------------------
r721
Inconsequential typo
---------------------
r720
Reverted version to 0.3a-WIP. Release version will be updated to 0.4a.
---------------------
r719
Added ability to load save state at startup using -load-state or InitStateFile (in config file)
---------------------
r718
work around for non working jtag code in star wars
---------------------
r717
Clamping must be applied before the spotlight is applied on step 1.0 h/w.
---------------------
r716
Fix step 1.0 games with NET_BOARD defined (Spindizzi)
---------------------
r715
Update network code (Spindizzi)
---------------------
r714
white space
---------------------
r713
Support both types of lightgun in lost world. 2nd type fixes some missing fx. Thanks to Any for pointing this out. To use the old type change analog_gun1 to gun1 in the xml file.
---------------------
r712
fix a few compile warnings
---------------------
r711
Add the netboard stuff to the ini file. Stops crashes when they are missing.
---------------------
r710
A hack to get spikeout to work (uncomment to enable) (Spindizzi)
---------------------
r709
Disable the netboard if EmulateNet=0 (Spindizzi)
---------------------
r708
remove extra qualifier
---------------------
r707
Update make files and fix build for the various options (Harry Tuttle)
---------------------
r706
fix header files
---------------------
r705
Remove using namespace std from the header files ..
---------------------
r704
Parse out line of sight position. Only known to be used by scud.
---------------------
r703
make the logic a bit clearer
---------------------
r702
fix Ian's bad maths :)
---------------------
r701
Update microtexture coordinate logic (Harry Tuttle)
---------------------
r700
fix illegal xml character sequence
---------------------
r699
simplify maths
---------------------
r698
comment out unneeded code
---------------------
r697
top/bottom were swapped for off axis projection
---------------------
r696
Rewrite projection maths based upon previously unknown viewport values. The previous values used roughly worked as the normals for frustum planes. Perhaps they were only used for culling and not actually rendering, as sometimes the values don't work correctly.
---------------------
r695
fix build
---------------------
r694
add the rest of the network board code (Spindizzi)
---------------------
r693
use the correct c++ deleters
---------------------
r692
use a few macros to make debugging less painful
---------------------
r691
Add network board emulation. Currently not linked with the rest of the project.
---------------------
r690
remove using namespace from headers ..
---------------------
r689
rename to avoid clash with a #define
---------------------
r688
add asynchronous sending functions
---------------------
r687
remove dead variable
---------------------
r686
Add basic networking files to supermodel. Currently these are win32 specific but can port to SDL or whatever later.
---------------------
r685
Fix the smallest texture lods (Harry Tuttle)
---------------------
r684
fix array out of bounds
---------------------
r683
remove debug code
---------------------
r682
mark as static
---------------------
r681
Implement sub 8x8 tile encoding. Fixes incomplete mipmap chains which previously only went down to 8x8 pixels before. It wasn't known these textures existed before. (Harry Tuttle)
---------------------
r680
Fix some bad/missing textures in fighting vipers. The game is referencing totally illegal texture sizes. The real h/w was just handling it somehow, whilst we were returning null for the textures. This might produce junk textures for the lower mipmaps, but this can be clamped in a later update.
---------------------
r679
Shift fog maths to fragment shader to fix bug in virtua on.
---------------------
r678
The vertex shader is run before clipping is done. If fogging values are calculated and clamped in the vertex shader it can interpolate bad values, since the range has been truncated. This happens if the polys overlap the near plane. The solution to clamp the values in the fragment shader. This fixes a bunch of fogging errors I long thought were transparency related errors in the ocean hunter.
---------------------
r677
Fixed shading also works with untextured polys. Fixes some shading issues in La machineguns. (Harry Tuttle)
---------------------
r676
Update step 1.5 luminous logic with Harry's findings. Fixes various shading issues.
---------------------
r675
Finish the hash function (Harry Tuttle)
---------------------
r674
Remove fixed shading check from specular.
---------------------
r673
Makefile fixes: Moved Windows-specific source files out of Makefile.inc, using 'bin' and 'obj' when bits are not explicitly set, and changed output file to Supermodel rather than Supermodel.exe, which works on all platforms.
---------------------
r672
Star wars is the only game to pass unsigned fixed shaded values (per vertex brightness) to the renderer. Originally we thought that the specular flag would turn on unsigned values since it's the only game to set specular with these polys, but this logic turned out to be incorrect. The JTAG interface seems to config the GPU to turn on this functionality. (Harry Tuttle)
---------------------
r671
update visual studio project
---------------------
r670
update spotlight code (Harry Tuttle)
---------------------
r669
Harry's patch: add dirtdvlsj and change "Star Wars Trilogy" to "Star Wars Trilogy Arcade"
---------------------
r668
New, saner Makefiles. They still require some customization but share a core Makefile include with auto-generation of dependencies. Multi-threaded building (e.g., make -j6) is supported now.
---------------------
r667
add missing header
---------------------
r666
Get rid of unused variable warning
---------------------
r665
Fix the sky in one of the levels in virtua on. The problem was caused by the fact we didn't split the mesh by the fog % burn through, so all the polys had the same value. We could use a vertex attrib, but % burn through fog is so rarely used and when it is, normally every poly in the mesh has the same value.
---------------------
r664
Fix tiny memory leak
---------------------
r663
---------------------
r662
update project files
---------------------
r661
work around for visual studio bug
---------------------
r660
fix missing header
---------------------
r659
cannot specify explicit initializer for arrays - fix for visual studio
---------------------
r658
- New work-in-progress frame timing code (disabled by default, compile with NEW_FRAME_TIMING defined to activate it)
- New JTAG emulation, moved into its own class, CJTAG
- Removed game-specific sun clamp hacks from CNew3D (JTAG and Real3D emulation will call the appropriate method to configure this at run-time)
- Removed JTAG from Real3D save state data and reused some of that space for new state variables having to do with the internal JTAG-based config as well as new frame timing state variables
---------------------
r657
Added functions to safely read and write bools
---------------------
r656
Added functions to extract bits as integer values and updated comment about bit vector layout
---------------------
r655
Added SetOnes() and SetZeros()
---------------------
r654
Safe guard against shifting empty register.
---------------------
r653
Hex formatting support extended to 64 bits.
---------------------
r652
Added ShiftOutLeft() and ShiftOutRight()
---------------------
r651
Renamed shift in/out functions to add/remove to be less ambiguous about their functionality
---------------------
r650
Added ToHexString() and made ostream serialization use this
---------------------
r649
Fix for VS2015 and newer by MetalliC
---------------------
r648
Added BitRegister class (for use in upcoming JTAG refactor): an inefficient but flexible container for manipulating a dynamic bit stream.
---------------------
r647
Calculating the length of the vertex results in quite broken values for fogging when the vertices are traversing the view frustum. The hardware also wouldn't have been doing a square root per vertex for fogging. Simply using the z values is enough. Todo check clamping of values for the near/far planes.
---------------------
r646
add interface for sun clamp
---------------------
r645
Optimise vertex data to cut down on unnecessary copying
---------------------
r644
Modern hardware does backface culling in window space by calculating the face normal for the polygon, then doing a dot product against the view vector. The real3d pro-1000 on the other hand passes a pre-calculated face normal for each polygon which is used for culling. We were using this face normal to rewind the polygons so that regular backface culling would work. This worked 99.9% of the time. However this was failing on some models in Virtua Striker. The reason was because the pre-calculated face normals being passed were actually completely different to the actual face normals for the poly (not just inverted like you would expect). This broke our code. The solution was to emulate face culling directly in the vertex shader using the pre-calculated face normals directly. Only minimally tested this but hopefully there are no obvious regressions.
---------------------
r643
DMA and register space access for Real3D have been unified and CReal3D::ReadRegister() returns little endian data now that the PowerPC handler byte-swaps back, to be consistent with how DMA registers work.
---------------------
r642
Changed fixed shaded logic for step 1.5 hw based upon Harry's findings
---------------------
r641
cosmetic
---------------------
r640
Harry made some important discoveries with regards to fixed shading on the model 3 (per vertex poly brightness values). Firstly values are allowed to be negative, and they are used as a drop in replacement in the standard lighting equation for the normal dot light vector. This quite radically changes the brightness in LA Machine guns, but now correctly matches the arcade.
---------------------
r639
Patched Sega Rally 2 DX to boot up. Enjoy Sega Rall 2 without MPEG music at 50% speed ;)
---------------------
r638
Fixed up memory region dumping code to use the new Util::FlipEndian functions
---------------------
r637
Fixed a bug in MCRFS instruction related to the VXSOFT, VXSQRT, and VXCVI bits. Thanks to Spindizzi.
---------------------
r636
Finish fixed shading for 2.0 hardware. I'm pretty sure it actually works identically to step 1.5 hardware. The oddball is LA machine guns where the viewport ambient doesn't seem to effect the brightness. But the ambient works differently in this game because it uses the unclamped light model. Still need to investigate if the diffuse factor effects fixed shading.
---------------------
r635
With fixed shading, when lighting is disabled, fixed shading appears to be a flat shaded version based upon the poly colour, instead of per vertex attributes.
---------------------
r634
Pad struct and align on 4 byte boundary.
---------------------
r633
simplify
---------------------
r632
fixed shading doesn't effect the alpha channel
---------------------
r631
Fixed shading (per vertex poly colours) on step 1.5 hardware have the viewport ambient value added to them. This fixes various shading on scud. To do this had to switch the maths to the vertex shader.
---------------------
r630
Fix attribute locations. glBindAttribLocation must be called linking
---------------------
r629
Added support for loading shaders from files for new engine and tile renderer. Changed help text to reflect new engine being default.
---------------------
r628
make the new3d engine default :] Legacy engine can still be selected using -legacy3d in the command line
---------------------
r627
use generic vertex attributes
---------------------
r626
Technically matrix casts are only allowed in glsl 1.2 and later
---------------------
r625
Remove debug code
---------------------
r624
Sometime ago I managed to work out that specular on the model3 is not real specular, and really is just an extension of diffuse lighting. But attempts were derailed by corner cases and the fact we were not handling the normals correctly. Anyway Harry managed to successfully come up with an algorithm, and coefficients that give an almost perfect match to specular on the model3, based soley on observations from video footage! He also worked out that the lighting on hardware 1.5 onwards appears to be unclamped (ie greater than 1). This quite radically changes the brightness of some of the games, but much better matches the original hardware.
---------------------
r623
cosmetic
---------------------
r622
Add the unclamped light model we know exists to the shaders.
---------------------
r621
Ski champ for some reason is passing denormalised numbers for the modelscale. This is causing a NaN in our shader killing the lighting for these models, so we simply skip these, since they are essentially zero anyway. Thanks to Harry for finding this bug :)
---------------------
r620
We were force normalising the result of matrix * normal, which looked correct in most cases. But this didn't preserve the scaling of the matrix, or the scaling of the model normals which resulted in many over bright areas. On it's own this generally worked, but games like Star Wars looked quite broken. Harry correctly figured out if you scale these normals by the scaling value that is sometimes present in the culling nodes the lighting looks correct. Still more work to do to correctly figure out the model3's lighting model.
---------------------
r619
Implement unclamped light model based upon Harry's findings. Unknown how this is turned on/off.
---------------------
r618
Use c++11 raw string literals
---------------------
r617
Clamp the ambient light to a max of 0.75. LA Machine guns seems to use an ambient value of 1, which is full bright for everything. But for some reason the hardware seems to treat this as ~ 0.75. This fixes various missing shading in the game. This doesn't seem to have any negative effects on other games I have tested. Clamping in the shader as we might need the full range of values for fixed shading.
---------------------
r616
remove WIP code
---------------------
r615
convert sun angle coordinate system outside of the shader
---------------------
r614
A second go at fixing the fixed lighting in star wars .. Fixed shading is per vertex poly colours. For some reason in star wars they are treated as unsigned values instead of signed, like in every other game. These polys are all marked with specular enabled, where as in the rest of the games they are missing this flag. That's the only difference I can find.
---------------------
r613
fix double offset
---------------------
r612
Revert previous fixed shading changes. Implement new logic for fixed shading on step 1.5 hardware based upon Harry's findings.
---------------------
r611
remove debug code
---------------------
r610
Fix the fix shading in star wars. Fixed shading still needs some work for mag truck and dirt devils to match the arcade.
---------------------
r609
cosmetic
---------------------
r608
Tweak scroll fog logic again .. should fix missing 2d fogging in Spikeout
---------------------
r607
rewrite scroll fog logic
---------------------
r606
Remove debug code
---------------------
r605
minor culling node optimisations
---------------------
r604
don't need cast, conversion is implicit
---------------------
r603
better state handling
---------------------
r602
parse out some culling node values
---------------------
r601
There are 2 bits in the polyheader labelled discard. When both are set ie in sega rally, the polygons are totally discarded. Scud seems to set bit 1 which seems halve the opacity.
---------------------
r600
The real3d has a flag that allows the hw to skip polygons so that they aren't rendered. Sega rally2 uses this for the dust effects. Triangles or quads that have this flag are junk, and look like random corruption if you try to draw them. We assumed polys that shared vertices with these bad polygons were also bad. It turns out the last few shared vertices in these polys were in fact valid, and thus the entire quad/triangle strip should be drawn and not discarded.
---------------------
r599
Linux Makefile fixed by Sergio Benjamin
---------------------
r598
light parameters must be clamped (HarryTuttle)
---------------------
r597
Virtua fighter has a few massive polys that shouldn't be visible. It looks like these polys have a bad matrix, but the node distances are all zero. This means the bounding box that encapsulates them is of size zero. Skipping over these entirely seems to fix the issue.
---------------------
r596
Updated Util::Config::Node unit tests with multiple leaf node test
---------------------
r595
remove unused variable
---------------------
r594
Extended -gfx-state to produce culling node analysis
---------------------
r593
Fixed behavior of Util::Config::Node when adding nested nodes (only leaf level can be duplicated; e.g., foo/bar and foo/bar will create one foo with two bar children, rather than two foo each with one bar).
---------------------
r592
Spindizzi's SCSP DSP fix for the VF3TB cave stage
---------------------
r591
Hooked up the command line help options (thanks, Spindizzi!) and updated copyright info
---------------------
r590
Removed unnecessary patches from lemans24
---------------------
r589
add missing include
---------------------
r588
Removed patches from source code and moved them into game XML file. Created ROMSet.cpp. Print a more descriptive error when game XML file fails to load.
---------------------
r587
Updated help text to include -game-xml-file option
---------------------
r586
Moved ROM patches into Games.xml and added patches to swtrilgy to make test emu work and to eca/ecau to pass the network check
---------------------
r585
add missing include
---------------------
r584
Harry Tuttle's fixes to fishing controls: invert fishing rod y and fishing stick y axis values so setting a negative axis is not necessary. Also changed reel to an analog input (similar to a pedal), whose neutral position is 0.
---------------------
r583
Harry Tuttle's change to Harley gear shift to use a single bit instead of two (no detectable difference in behavior, though)
---------------------
r582
Removed unused file: Game.cpp
---------------------
r581
Fixed lamachin title -- thanks Jiterdomer!
---------------------
r580
cosmetic
---------------------
r579
Daytona seems to use this completely undocumented feature of the real3d pro-1000, the ability to invert texture colours. This patch fixes the colours on the cars, and the signs which should flash by alternating their colours.
---------------------
r578
Removed last vestiges of CINIFile from input system and removed Games.cpp, INIFile.cpp, and ROMLoad.cpp
---------------------
r577
Using Game::INPUT_* instead of obsolete GAME_INPUT_*
---------------------
r576
Changes by Harry Tuttle: game list more accurately reflects MAME's, added a root level XML "games" node, print ROM version along with title
---------------------
r575
Fixed graphics state analyzer
---------------------
r574
Added cstring
---------------------
r573
Added remaining games largely thanks to the efforts of the mysterious '阳阳'
---------------------
r572
more whitespace ..
---------------------
r571
white space ..
---------------------
r570
Cache dereferenced values. Cuts loading time from 15 to 5 seconds in debug mode :)
---------------------
r569
Included cstring for memcpy
---------------------
r568
Rewrote large parts of ROM loading code to more gracefully handle merged sets where overwritten parent ROMs are not present. Better handling of multiple zip files. Optimized ROM loading when stride and chunk size are equal and got rid of safety checks that I don't believe are necessary any longer.
---------------------
r567
Rewrite the spot light code, and implement the missing fog logic. (HarryTuttle)
---------------------
r566
Fixed missing file detection (set_difference requires inputs sorted by desired key) and improved error reporting
---------------------
r565
Made scuda a child set of scud
---------------------
r564
Removed TODO comment
---------------------
r563
Smarter game selection when a zip file contains multiple games (prefer child when parent is also present in zip)
---------------------
r562
Warn about parent ROM sets having parent defined (not allowed)
---------------------
r561
Print missing files from games
---------------------
r560
Support loading of parent ROM sets (e.g., scudp will load scud)
---------------------
r559
compile fixes
---------------------
r558
MSYS2 compile fix (HarryTuttle)
---------------------
r557
Replace glsl 3 function. Apparently this doesn't work on OS X.
---------------------
r556
fix compile warnings
---------------------
r555
Sort games found in zip file by descending order of matching files so that game with the *most* matching files is picked
---------------------
r554
Added: bass, daytona2, dirtdvls, harley, lemans24, scuda, skichamp, spikeofe, spikeout, von2, vs2, vs298, vs2v991
---------------------
r553
Fixed scud mpeg roms
---------------------
r552
don't ask
---------------------
r551
fix project files
---------------------
r550
fix multiple default constructors
---------------------
r549
add missing include
---------------------
r548
Massive internal change: removed CConfig object and replaced it with a hierarchical config object system (Util::Config::Node). Games are now defined in an XML file. Hopefully I didn't break too many things :/
---------------------
r547
Cosmetic: renamed step to stepping
---------------------
r546
No need to enclose values in quotes when writing INI file. Print that file was saved to console and log.
---------------------
r545
Added a TODO list
---------------------
r544
Added srally2
---------------------
r543
Encapsulated zip file handling and added a "parent" member to the Game struct
---------------------
r542
Checking in GameLoader.cpp, GameLoader.h, and ROMSet.h (not currently used in build) because I need to make a substantial modification in my local tree
---------------------
r541
Fixed dayto2pe sample ROMs
---------------------
r540
The mipmap data in some games seems to being sent behind the base textures, and when the base textures are bound and created the mipmap chain is not yet valid in memory. To fix this, we invalidate the base textures, when the mipmap chain is updated.
---------------------
r539
Just use 1 hash map for the textures, instead of 1 for each format. Should greatly speed up texture invalidation. TODO invalidate texture mipmaps .. the h/w actually seems to stream the mipmap data separate from the base textures in some cases.
---------------------
r538
cosmetic
---------------------
r537
calculate the correct number of mipmap textures
---------------------
r536
Originally to do mipmapping, I just used the base texture and then let the h/w automatically generate the mipmap chain. I wasn't sure the real3d api even exposed to the programmer the ability to directly set the mipmap data. Anyway, as pointed out by HarryTuttle scud actually uses darker mipmaps for an effect on the water. This patch uses the uses the actual mipmap data from the real3d memory. Only tested scud, so hopefully I didn't cause any cataclysmic errors :)
---------------------
r535
Added fvipers2
---------------------
r534
use 32bit types
---------------------
r533
Added more games to Games.xml: dayto2pe, eca, lamachin, lostwsga, magtruck, oceanhun, vf3. Fixed swtrilgy (MPEG music).
---------------------
r532
Added 8- and 16-bit tilegen VRAM access handlers to fix missing columns in Star Wars Trilogy tilemaps. Thanks to Spindizzi for initial confirmation that this was the problem.
---------------------
r531
fix scissor code for native resolution (thanks HarryTuttle)
---------------------
r530
remove debug code
---------------------
r529
The model3 has this weird issue where the 2d and 3d planes are misaligned by 2 pixels. Originally I made a quick hack that shifted subviewports 2 pixels. But apparently that wasn't enough, as the issue effects everything. Thanks to HarryTuttle for pointing this out, and making a patch :)
---------------------
r528
Texture coordinates seem to be scaled relative to the fixed texture size of 256. Thanks to HarryTurtle for figuring this crazy logic out. The road in scud matches the arcade now almost perfectly.
---------------------
r527
Change microtexture scaling values. Seem to much better match expected output. (thanks HarryTuttle)
---------------------
r526
More fixes to microtexture coordinates. They go vertically up and down the texture sheet, and not horizontally as I had guessed. (Thanks HarryTurtle)
---------------------
r525
Fix microtexture texture coordinates (thanks HarryTuttle)
---------------------
r524
fix the bad projection matrix in dirt devils and lost world
---------------------
r523
The hardware actually seems to treat the quad primitive as a triangle strip. The winding order can actually change between the first and second 'triangle' inside the quad, leading our code to incorrectly backface cull the 2nd triangle. So we must re-calculate the winding order for the 2nd triangle. This fixes a massive missing poly in the road in la machine guns, probably other games too. The hardware is unaffected by the winding order, as it doesn't calculate the poly normal from the vertices directly, instead it uses the provided normal in the poly header.
---------------------
r522
Support high priority polygons. They are treated as a separate overlay plane by the h/w. Fixes some missing fx in harley.
---------------------
r521
Add high priority poly function. Used by harley.
---------------------
r520
When the translator map?? bit is enabled the colour format seems to be 4-5bit. Max observed value is 16. Scaling these values up seems to produce the correct colours in harley and dirt devils.
---------------------
r519
Put a check in there, in case near values come out really close to zero. Fixes a slight corner case in lemans24. A check probably should have been there originally :)
---------------------
r518
Dynamically calculate near/far frustum planes for each viewport priority layer by clipping the overlapping meshes with the 4 frustum planes. A bit brute forcy, but the results are great.
---------------------
r517
mark the nodes without valid distances set
---------------------
r516
cosmetics
---------------------
r515
only need to cull against 4 clipping planes
---------------------
r514
For scroll fog set the viewport to the entire area, allows fog to work in widescreen mode
---------------------
r513
properly cache un-normalised texture coordinates, fixes a texturing regression in Daytona
---------------------
r512
fix microtexture colours
---------------------
r511
Only check scroll fog value from the first viewport from each priority layer. Seems to fix problems with background fogging.
---------------------
r510
delete unused files
---------------------
r509
Support microtexture relative scale. The exact scaling values are not known, but a 2x difference between each level seems highly likely. We know from visual inspection from scud that the first level is exactly 4. The SDK has a function that accepts a float input for scale and does this. lod = (1/scale) + 0.5. If lod>3 lod = 3. So this is our best guess.
---------------------
r508
The model3 shares vertices between polygons. But it also seems to do this for polygons with different textures. This means the texture coordinate for potentially different sizes textures is shared. Currently this was broken leading to some corrupted looking textures. The fix is to simply cache the old texture coordinates, and recalculate the normalised coordinates in the new texture coordinate space.
---------------------
r507
refactor
---------------------
r506
tweak scroll fog logic
---------------------
r505
clean up
---------------------
r504
minor optimisations
---------------------
r503
Only check the geometry for culling with valid node distances. We will need this later.
---------------------
r502
Revert changes .. if we clip against the real3d clipping planes, our wide screen hack won't work anymore.
---------------------
r501
Get clipping planes directly from the viewport data. Also no need to clip against a far plane
---------------------
r500
Need to walk back in memory and check the previous texture tiles to see if they overlap with the current textures being invalidated. If they overlap they must be cleared too. Algorithm is a bit brute forcy and more expensive than I'd like but it fixes the bad textures in Spikeout.
---------------------
r499
Skip processing the mipmap section of memory. The real3d doesn't create mipmaps down to 1x1 that is needed to do mipmapping properly on modern hardware anyway.
---------------------
r498
Clear queued texture uploads after processing them. Saves re-uploading the same textures whilst paused
---------------------
r497
More accurate 2D colours. White should map to 255 not 248
---------------------
r496
Fix the shading in scud. When lighting is disabled the fixed shading intensities appear to be offset values added to the colour value.
---------------------
r495
change throttle logic for smoother playback, we could get closer to 60fps if SDL timers weren't so lame
---------------------
r494
change scroll fog logic to parse the lower priority viewports first
---------------------
r493
implement scroll fog, fixes a tonne of missing fx
---------------------
r492
Parse out scrollfog value used for blending over the 2d background.
---------------------
r491
Fix scenes with missing fog in ocean hunter. The game was passing negative fog densities. The sign seems to be ignored by the model 3.
---------------------
r490
same fix for legacy engine
---------------------
r489
our scratch mem must match correct max texture size
---------------------
r488
Fix the sky in advanced stage in Dirt Devils. It uses a massive (for 1998 :p) 1024x256 texture.
---------------------
r487
Support multiple microTextures, fixes crazy texturing in sega ski champ. I'm assuming microtextures are always located on the other memory bank to the base texture. This logic seems to work for all our games anyway.
---------------------
r486
Added RGBA4 format to BMP file writer
---------------------
r485
Made parsing of hex values more robust
---------------------
r484
Initial check in of Games.xml (only two games defined so far)
---------------------
r483
Fixed new 3D engine's fragment shader to work on some older GPUs by reducing the number of discard statements to one
---------------------
r482
Updated game loader to make use of latest changes to config tree: generic value types and eliminated pointers
---------------------
r481
Support for hex numbers in string to integer conversions (with '0x' prefix) and special keywords in string to bool conversions (true/false/on/off/yes/no)
---------------------
r480
Added Util::Stricmp() and small tweak to serialization of Util::Format
---------------------
r479
Added support for multiple value types in config nodes.
Removed Ptr_t and ConstPtr_t from public interface, moved config tree builders into their own file, and made them pass by reference and value.
Exceptions can be thrown now on lookup failures.
Removed s_empty_node -- failed lookups in [] operator create a permanent "hidden" child.
Updated comment in NewConfig.cpp.
---------------------
r478
Added copy constructor and assignment operator using C++11 move semantics
---------------------
r477
Defined inputs
---------------------
r476
Support for deep copies via assignment operator. This will pave the way for elimination of Ptr_t and ConstPtr_t.
---------------------
r475
Added Write() method to write buffer to an ostream, and added an ostream operator overload
---------------------
r474
Changes to game loader interface, added 32-bit word swapping utility function
---------------------
r473
Added Logger.o to OSX and UNIX Makefiles. Should fix compilation on these platforms.
---------------------
r472
Removing MSVC Makefile
---------------------
r471
Initial check-in of game loader (not operational yet)
---------------------
r470
Byte swapping utility function
---------------------
r469
Added ValueAsBool(), ValueAsBoolWithDefault(), and ValueAsUnsignedWithDefault()
---------------------
r468
update visual studio project
---------------------
r467
Added Util::Config::Node::ValueAsUnsigned() with support for base 10 and 16 ('0x' prefix)
---------------------
r466
Added Util::ToLower()
---------------------
r465
Config::Node::Add() now supports nested keys (e.g., Add("foo/bar/baz", "0") will create three nested config nodes)
---------------------
r464
Changed Config::Node::Create() to Add() and added a Set() method (for INI semantics). Turned Test_Config.cpp into a proper unit test reporting pass/fail for each test.
---------------------
r463
Adding tinyxml2 lib
---------------------
r462
XML parsing support for new config system
---------------------
r461
Moved logging functions into OSD/Logger.cpp (eventually logging system will be reworked entirely) and added new config tree structure (not yet used).
---------------------
r460
Updated Win32 gcc, UNIX, and OSX Makefiles
---------------------
r459
Updated Fragment_NoSpotlight.glsl (fragment shader for legacy engine without spot light)
---------------------
r458
disable spot light if the extent is zero
---------------------
r457
Saving progress on scudp disassembly
---------------------
r456
parse some unimplemented spot light values from viewport
---------------------
r455
Added disassembly of Scud Race Plus
---------------------
r454
Marked offset of code in CROM for scudp
---------------------
r453
stencil translucent non textured polys
---------------------
r452
refactor
---------------------
r451
Cull geometry based on the culling distances in the nodes. Should give some sort of speedup on lower end pcs where the CPU is predominantly the bottle neck.
---------------------
r450
Fix compilation when SUPERMODEL_DEBUGGER enabled
---------------------
r449
cache bound texture
---------------------
r448
cleaning
---------------------
r447
Added drive board and lamp raw outputs (submitted by SailorSat)
---------------------
r446
support negative exponent
---------------------
r445
Add code to convert the 16bit float format used by the hardware back into float 754 format. 16bit floats are used for the culling radiuses in the nodes.
---------------------
r444
Added -vsync and -no-vsync option descriptions to command line help
---------------------
r443
remove dead code
---------------------
r442
Added a Split() function to Util::Format.
---------------------
r441
XInput fix: search for all three possible XInput DLLs (xinput1_4.dll, xinput1_3.dll, and xinput9_1_0.dll). I have not tested 32-bit builds.
---------------------
r440
Check for overflow. Fixes some fade effects
---------------------
r439
Legacy engine: fixed texture offsets (VON2 was broken by last commit). Texture coordinates are wrapped within the bank first and then the bank can be swapped. This also allows the full range of texture coordinate bits to be used as specified in the Pro-1000 SDK.
---------------------
r438
Y texture offsets are just 5 bits. Texture coordinate wrap around happens in the same texture sheet, not into the next one. Fixes various texturing bugs in sega rally.
---------------------
r437
Legacy engine: Culling node texture Y offset is only 5 bits (fixes Sega Rally 2 initials decal)
---------------------
r436
New engine: modf -> std::modf (fixes compilation on gcc)
---------------------
r435
Legacy engine: expanded color index to 12 bits from 11. Does this affect any games?
---------------------
r434
New engine: expanded color indices to 12 bits from 11. Added PolygonHeader::ColorIndex and SensorColorIndex methods.
---------------------
r433
Forgot to describe one polygon header field in Model.cpp
---------------------
r432
Updated polygon header descriptions
---------------------
r431
Legacy engine: Added stencil buffering for layered models and shadows. Not as accurate as new engine because stencil buffering is applied per-model rather than per-polygon, so the entire model must consist of stencil-tested polygons. In practice, though, this seems to work fine.
---------------------
r430
Optimise non smooth texture repeat to only effect texture boundaries, rather than poly boundaries. Fixes a few ugly artefacts
---------------------
r429
fix transparency values, values are from 0-32, not 0-16 where the top bit disables
---------------------
r428 - Some Sauce by Ian.
Stencil layered polys, fixes shadow z fighting in some of the games. Other games maybe using stipple alpha (yuck), not sure yet.
---------------------
r427
remove bad rom patches
---------------------
r426
don't clamp after spot light code, looks much closer to h/w
---------------------
r425
Hooked up Sega Bass Fishing / Get Bass controls
---------------------
r424
basic spotlight code
---------------------
r423
Fixed fog in legacy 3D engine. Now computed same as new engine, with 4-bit light modifier and corrected fog mixing in shaders.
---------------------
r422
Updated legacy engine vertex shader to use specular shininess and coefficient along with old equation. Still not correct (too strong compared to the old method).
---------------------
r421
fix alpha testing for texture format 7
---------------------
r420
remove dead code
---------------------
r419
support 4 bit textures
---------------------
r418
correctly mask pointers
---------------------
r417
optimise hash table access for better performance
---------------------
r416
support colour table pointer
---------------------
r415
remove debug code
---------------------
r414
Fog start is apparently the fog value at the viewport (0,0,0) I guess like an ambient value. Default value is negative
---------------------
r413
Non luminous polygons seem to be effected by the lightModifier also. API seems to have no restriction in this regard, neither does the SDK. The ocean hunter requires this for some of the 2d stuff. Light modifier values only seem to exist between 0-16, which is strange for 5 bits of precision. Possibly top bit disables the effect and lower 4 bits are the value. This clearly differs from the SDK, but without this the values come out clearly wrong.
---------------------
r412
adjust (hopefully correct) fog equation
---------------------
r411
preliminary specular
---------------------
r410
- Removed a stale comment
The read-me instruction for the custom MP2 file is in the zip file.
Note: This build had netboard emulation built-in.
Here are the SCSP options in the command line:
-new-scsp (default, uses Conversus' SCSP code)
-legacy-scsp (uses old Supermodel SCSP code)
Arguments regarding Speakers:
-channels=4 (Quadrophonic Speakers)
-channels=2 (Stereo Speakers)
True Refresh Rate Argument (enables 57.5hz):
-true-hz
Here's an option to dump textures in Model 3 Games:
-dump-textures
