render96/render96ex

Fork of https://github.com/sm64-port/sm64-port with additional features.

render96ex's Issues

Tester Branches: Anomalies in the sound implementation

Describe the bug
The sound behavior in the testing branch for this port does not match the original game. Examples include certain sound effects being missing or music playing at times where it normally wouldn't.

To Reproduce
Steps to reproduce the behavior:

1. Build the tester or tester_rt64alpha branch with SM64PCBuilder2
2. Start the game
3. Load a save file, then listen as Mario says "Okie-Dokie" without the accompanying star sound
4. Enter a level / select a star
5. Listen as Mario says "Lets-a-go" without the accompanying star sound

Additional examples:
Pause the game when inside a level, such as Bob-omb Battlefield. The music will continue on instead of pausing

Expected behavior
In the master branch of this port and in other SM64 ports, the audio behavior is unchanged from the original game. The testing branch of this port has some missing sound effects and music playback adjustments, and I'm not sure if these changes are by design or just long-standing bugs in the port. Being on the master branch, you sacrifice over a years worth of other improvements

Desktop (please complete the following information):

• OS: Windows 11 Pro 21H2
• Browser: Google Chrome
• Version: 107.0.5304.88
• CPU: AMD Ryzen 9 5900X
• GPU: NVIDIA GeForce RTX 3060 Ti
• RAM: 64 GB DDR4

Original models?

How do I access the original models instead of these weird low poly ones that come with this project? They don't seem to be the original ones.

3D Trees Face the Camera

Describe the bug
With Render96 v2.2 models enabled or VaniLla96 1.0.1 MP enabled the trees will always rotate to face the camera as though they are 2D despite a 3D model being enabled.

Expected behavior
Trees do not rotate to face the camera when Render96 v2.2 or VaniLla96 1.0.1 are enabled.

Details:

• Windows 10
• Render96ex v2.2 alpha build
• Odyssey Moveset v5.1.2

Additional context
It is worth noting in older builds of Render96 before DynOs was implemented fully that 3D trees did not face the camera yet they do in the current versions. It is also worth noting that other 3D assets rotate as well, e.g. the 1-Up mushroom, though it doesn't really matter if those rotate or not since they aren't as intrusive as the trees.

Issues with the new "transparency GL" option.

Issue:
I have been testing with this new option, it fixes the problems with transparent models but it introduces new issues (at least in my system) constant flickering with shadows, lighting texture glitches, etc.

To Reproduce:
A quick and easy way to reproduce this problem is to go to the main hall in the castle and see how Mario's shadow blinks. Another method is to go to Course 5 "Big Boo's Haunt" and see how the "halos" produced by the torches constantly flicker depending on the camera angle.

Video:
https://www.youtube.com/watch?v=rI76VKqH7NY

Timestamps: 0:22 - 0:39 - 0:54 & 1:13

• OS: Windows 10 21H2
• GPU: Geforce GTX970 @ GTX 511.79

Rom Hacks possible?

Was wondering if it is possible to use a RomHack instead of the Normal Rom. Get a Issue with it. Also it is Extended (24MB)

Error within the readme

Describe the bug
The readme in the master branch describes the process of installing the Models and Textures with EXTERNAL_DATA and DynOS, however neither are present after building the master branch.

To Reproduce
Steps to reproduce the behavior:

1. Clone the master branch
2. Build the master branch
3. The folders res and dynos are not present.

Expected behavior
The folders "res" and "dynos" exist, allowing to put the files within them.

Desktop:

• OS: Arch Linux
• Kernel: 6.1.39-3-lts

Additional context
If I am the only one having this problem, it might be caused by the 60 FPS patch, or the TEXTURE_FIX Build Flag.

Suggestion: add your own cheats, save states, and frame advance

JP build is broken

First build has this:
tools/aiff_extract_codebook sound/samples/course_start/00_la.aiff >build/jp_pc/sound/samples/course_start/00_la.table
tools/aiff_extract_codebook sound/samples/instruments/00.aiff >build/jp_pc/sound/samples/instruments/00.table
Traceback (most recent call last):
File "/home/Aaron Luke/Render96ex/tools/demo_data_converter.py", line 78, in
main()
File "/home/Aaron Luke/Render96ex/tools/demo_data_converter.py", line 60, in main
with open("assets/demos/" + item["name"] + ".bin", "rb") as file:
FileNotFoundError: [Errno 2] No such file or directory: 'assets/demos/bitdw.bin'
make: *** [Makefile:750: build/jp_pc/assets/demo_data.c] Error 1
make: *** Deleting file 'build/jp_pc/assets/demo_data.c'

Running US build first will give you:

python3 tools/assemble_sound.py --sequences build/jp_pc/sound/sequences.bin build/jp_pc/sound/bank_sets sound/sound_banks/ sound/sequences.json sound/sequences/jp/01_cutscene_collect_star.m64 sound/sequences/jp/02_menu_title_screen.m64 sound/sequences/jp/03_level_grass.m64 sound/sequences/jp/04_level_inside_castle.m64 sound/sequences/jp/05_level_water.m64 sound/sequences/jp/06_level_hot.m64 sound/sequences/jp/07_level_boss_koopa.m64 sound/sequences/jp/08_level_snow.m64 sound/sequences/jp/09_level_slide.m64 sound/sequences/jp/0A_level_spooky.m64 sound/sequences/jp/0B_event_piranha_plant.m64 sound/sequences/jp/0C_level_underground.m64 sound/sequences/jp/0D_menu_star_select.m64 sound/sequences/jp/0E_event_powerup.m64 sound/sequences/jp/0F_event_metal_cap.m64 sound/sequences/jp/10_event_koopa_message.m64 sound/sequences/jp/11_level_koopa_road.m64 sound/sequences/jp/12_event_high_score.m64 sound/sequences/jp/13_event_merry_go_round.m64 sound/sequences/jp/14_event_race.m64 sound/sequences/jp/15_cutscene_star_spawn.m64 sound/sequences/jp/16_event_boss.m64 sound/sequences/jp/17_cutscene_collect_key.m64 sound/sequences/jp/18_event_endless_stairs.m64 sound/sequences/jp/19_level_boss_koopa_final.m64 sound/sequences/jp/1A_cutscene_credits.m64 sound/sequences/jp/1B_event_solve_puzzle.m64 sound/sequences/jp/1C_event_toad_message.m64 sound/sequences/jp/1D_event_peach_message.m64 sound/sequences/jp/1E_cutscene_intro.m64 sound/sequences/jp/1F_cutscene_victory.m64 sound/sequences/jp/20_cutscene_ending.m64 sound/sequences/jp/21_menu_file_select.m64 build/jp_pc/sound/sequences/00_sound_player.m64 -DVERSION_US -D_LANGUAGE_C -DGIT_HASH=""f5735aa8"" -DNON_MATCHING -DAVOID_UB $(cat build/jp_pc/endian-and-bitwidth)
sequences.json assigns sound banks to 22_cutscene_lakitu, but there is no such sequence file. Either remove the entry (or set it to null), or create sound/sequences/22_cutscene_lakitu.m64.
make: *** [Makefile:711: build/jp_pc/sound/sequences.bin] Error 1

Sm 64

Cool Cool Mountain Crashes When Entering Painting

Cool Cool Mountain Crashes When Entering PC Version

(Render96rt) Star is flat

For some reason the star model in the DynOS pack doesn't have any RTX effects, it's a simple flat texture.

Green Keys not saving to/loading from save file correctly on Linux/macOS

Describe the bug
So it's a bit tricky to explain, but whenever I collect one of the green keys, it does not save correctly. It'll save the first key from Bobomb Battlefield, but no others. even if the first key I got wasn't from BoB.

If I edit the save file after compiling with "text saves", I can add in the amount of 1's needed to obtain all the keys, but saving a star or anything after that reverts it, where the save file only says 1, rather than how many zeroes.

Everything else saves just find, stars and caps, and the unlockable the keys unlock saves if I happen to save while that's active, but switching back requires editing the save or reobtaining the keys again.

This was tested on the previous stable 1.4.2, but within the SM64 Discord I and one other user can confirm this issue persists into the current 2.0 Alpha. We're not sure if this is a Linux only issue or what it is.

To Reproduce
Steps to reproduce the behavior:

1. Go to any level with one of the keys, Bob Omb Battlefield and the Secret Slide have two quick and easy ones.
2. Pick up the key
3. Quit the game, the keys autosave but saving with a star is optional
4. Reload the game

Expected behavior
You should have 2 Keys

Actual behavior
You only have 1 key, regardless of the amount you had prior

Desktop (please complete the following information):

• Manjaro Linux 5.9.1.1
• Version Super Mario 64 Render96ex 1.4.2

Additional context
My version was built with the smlinux scripts, tested various configurations, including a vanilla Render96ex with no extra patches or cheats, and this issue persists within all of them.

Building on FreeBSD

I'm trying to build this for FreeBSD in hopes of a PR (which i submitted for sm64pc/sm64ex), but the linker complains about an undefined symbol: stderr (output when executing gmake:

gmake: Entering directory '/usr/home/ezntek/Sources/Render96ex/tools'
cc -I../include -I. -Wall -Wextra -Wno-unused-parameter -pedantic -std=c99 -O2 -s -Iaudiofile -Wno-uninitialized sdk-tools/tabledesign/codebook.c sdk-tools/tabledesign/estimate.c sdk-tools/tabledesign/print.c sdk-tools/tabledesign/tabledesign.c -o tabledesign -lm -Laudiofile -laudiofile -lstdc++ -lm
ld: error: undefined symbol: stderr
>>> referenced by audiofile.cpp
>>>               audiofile.o:(defaultErrorFunction(long, char const*)) in archive audiofile/libaudiofile.a
>>> referenced by audiofile.cpp
>>>               audiofile.o:(defaultErrorFunction(long, char const*)) in archive audiofile/libaudiofile.a
>>> referenced by audiofile.cpp
>>>               audiofile.o:(defaultErrorFunction(long, char const*)) in archive audiofile/libaudiofile.a
>>> referenced 13 more times

ld: error: undefined symbol: std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >::_M_create(unsigned long&, unsigned long)
>>> referenced by audiofile.cpp
>>>               audiofile.o:(void std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >::_M_construct<char const*>(char const*, char const*, std::forward_iterator_tag) (.isra.0)) in archive audiofile/libaudiofile.a

ld: error: undefined symbol: __errno_location
>>> referenced by audiofile.cpp
>>>               audiofile.o:(FileModule::reportWriteError(long, long)) in archive audiofile/libaudiofile.a

ld: error: undefined symbol: stdout
>>> referenced by audiofile.cpp
>>>               audiofile.o:(_af_print_filehandle(_AFfilehandle*)) in archive audiofile/libaudiofile.a

ld: error: undefined symbol: std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >::_M_append(char const*, unsigned long)
>>> referenced by audiofile.cpp
>>>               audiofile.o:(AudioFormat::description[abi:cxx11]() const) in archive audiofile/libaudiofile.a
>>> referenced by audiofile.cpp
>>>               audiofile.o:(AudioFormat::description[abi:cxx11]() const) in archive audiofile/libaudiofile.a
>>> referenced by audiofile.cpp
>>>               audiofile.o:(AudioFormat::description[abi:cxx11]() const) in archive audiofile/libaudiofile.a
>>> referenced 4 more times

ld: error: undefined symbol: std::__throw_length_error(char const*)
>>> referenced by audiofile.cpp
>>>               audiofile.o:(AudioFormat::description[abi:cxx11]() const) in archive audiofile/libaudiofile.a
>>> referenced by audiofile.cpp
>>>               audiofile.o:(void std::vector<unsigned long, std::allocator<unsigned long> >::_M_realloc_insert<unsigned long const&>(__gnu_cxx::__normal_iterator<unsigned long*, std::vector<unsigned long, std::allocator<unsigned long> > >, unsigned long const&)) in archive audiofile/libaudiofile.a
>>> referenced by audiofile.cpp
>>>               audiofile.o:(void std::vector<SharedPtr<Chunk>, std::allocator<SharedPtr<Chunk> > >::_M_realloc_insert<SharedPtr<Chunk> >(__gnu_cxx::__normal_iterator<SharedPtr<Chunk>*, std::vector<SharedPtr<Chunk>, std::allocator<SharedPtr<Chunk> > > >, SharedPtr<Chunk>&&)) in archive audiofile/libaudiofile.a
>>> referenced 1 more times
cc: error: linker command failed with exit code 1 (use -v to see invocation)
gmake: *** [Makefile:47: tabledesign] Error 1
gmake: Leaving directory '/usr/home/ezntek/Sources/Render96ex/tools'
Makefile:236: *** Failed to build tools.  Stop.

)

setting CC=gcc in the makefile results in something similar:

gmake: Entering directory '/usr/home/ezntek/Sources/Render96ex/tools'
gcc -I../include -I. -Wall -Wextra -Wno-unused-parameter -pedantic -std=c99 -O2 -s -Iaudiofile -Wno-uninitialized sdk-tools/tabledesign/codebook.c sdk-tools/tabledesign/estimate.c sdk-tools/tabledesign/print.c sdk-tools/tabledesign/tabledesign.c -o tabledesign -lm -Laudiofile -laudiofile -lstdc++ -lm
/usr/local/bin/ld: audiofile/libaudiofile.a(audiofile.o): in function `defaultErrorFunction(long, char const*)':
audiofile.cpp:(.text+0x1130): undefined reference to `stderr'
/usr/local/bin/ld: audiofile.cpp:(.text+0x113c): undefined reference to `stderr'
/usr/local/bin/ld: audiofile.cpp:(.text+0x114b): undefined reference to `stderr'
/usr/local/bin/ld: audiofile/libaudiofile.a(audiofile.o): in function `Track::print()':
audiofile.cpp:(.text+0x4450): undefined reference to `stderr'
/usr/local/bin/ld: audiofile.cpp:(.text+0x446a): undefined reference to `stderr'
/usr/local/bin/ld: audiofile/libaudiofile.a(audiofile.o):audiofile.cpp:(.text+0x4486): more undefined references to `stderr' follow
/usr/local/bin/ld: audiofile/libaudiofile.a(audiofile.o): in function `FileModule::reportWriteError(long, long)':
audiofile.cpp:(.text+0x5749): undefined reference to `__errno_location'
/usr/local/bin/ld: audiofile/libaudiofile.a(audiofile.o): in function `_af_print_filehandle(_AFfilehandle*)':
audiofile.cpp:(.text+0x6b33): undefined reference to `stdout'
/usr/local/bin/ld: audiofile/libaudiofile.a(audiofile.o): in function `ModuleState::print()':
audiofile.cpp:(.text+0x9f55): undefined reference to `stderr'
/usr/local/bin/ld: audiofile.cpp:(.text+0x9fa4): undefined reference to `stderr'
/usr/local/bin/ld: audiofile.cpp:(.text+0x9fd5): undefined reference to `stderr'
/usr/local/bin/ld: audiofile.cpp:(.text+0xa02e): undefined reference to `stderr'
collect2: error: ld returned 1 exit status
gmake: *** [Makefile:47: tabledesign] Error 1
gmake: Leaving directory '/usr/home/ezntek/Sources/Render96ex/tools'
Makefile:236: *** Failed to build tools.  Stop.

Digging a bit deeper, I found this, which suggested that this is a feature in glibc only. Is this glibc requirement necessary to some features in R96ex (seeing as this compalins about libaudiofile and c++), or can this restriction be removed? Don't want to make any unfair comparisons, but sm64ex (which is what this is forked from) supports FreeBSD in terms of compilation. Any help would be appreciated, because I'm building a sm64pclauncher/smlinux inspired kind of thing for Unix-like systems.

1up textures glitch

Describe the bug
sometimes the 1up textures look weird from distances

To Reproduce
Steps to reproduce the behavior:

1. Move away from the 1up
2. See glitch

Expected behavior
Normal 1up textures

Screenshots

Super.Mario.64.Render96ex.OpenGL.08153516.2023-03-30.20-30-17.mp4

Desktop (Please fill the following information):

• Win11
• Browser Microsoft edge
• Mod version 3.2

tester_rt64alpha - Unicode letter â (00226) is never called and rendered properly on Peach's letter and Level name Pause screen

Describe the bug

Unicode letter â (00226) always renders as the previous character called offset to the bottom

Compiled tester_rt64alpha on Intel Mac and changed language to 10 (French), and the letter from Peach shows this bug right away

Changing the affected character in the french translation file to uppercase  {00194} for example shows the correct uppercase version of the character

(using graphic packs or not yields the same results)

This only affects Peach's letter and level name on the Pause screen as far as I can tell (not billboards)

To Reproduce

Compile tester_rt64alpha branch and change language to 10 (French)

Expected behavior

{"{00226}", {110, 0}}, // LATIN SMALL LETTER A WITH CIRCUMFLEX to be rendered as character_00E2.ia4.png

Screenshots

Desktop (please complete the following information):

macOS 11.7.1

Alpha and RT builds compile in German language by default

When I compile builds from the alpha and tester_rt64alpha branches, the language is always set to German and I can find no way to alter this fact. This is odd as I am using a US rom for the builds.

I am compiling the builds manually on PopOS.

I find I can alleviate this issue by copying the English language json files and replacing the German ones with them, although this process would need to be repeated for every compilation I do.

Breaking the wing cap box on Shifting Sand Land crashes when playing as Luigi (on the RTX build)

Describe the bug
You crash if you break the red wing cap box on Shifting Sand Land.

To Reproduce
Steps to reproduce the behavior:

1. Go to Shifting Sand Land
2. Break the wing cap box
3. Crash

Expected behavior
Not crash

Other info

• Win 11
• RTX build
• I have not tried this with other characters, it may crash with them too. I don't know.

Problem compiling on raspberrypi

I'm trying to compile the windows build on a raspberry pi.

I get the following error :

`
$ make -j4 WINDOWS_BUILD=1
make: Entering directory '/home/pi/supermario64/render64ex/Render96ex/tools'
make: Nothing to be done for 'default'.
make: Leaving directory '/home/pi/supermario64/render64ex/Render96ex/tools'
cc -c -O2 -I include -I build/us_pc -I build/us_pc/include -I src -I . -DDYNOS -DRAPI_GL=1 -DWAPI_SDL2=1 -DAAPI_SDL2=1 -DCAPI_SDL2=1 -DHAVE_SDL2=1 -I/usr/include/SDL2 -I/opt/vc/include -I/opt/vc/include/interface/vcos/pthreads -I/opt/vc/include/interface/vmcs_host/linux -D_REENTRANT -DVERSION_US -D_LANGUAGE_C -DGIT_HASH=""8e2eea10"" -DNON_MATCHING -DAVOID_UB -DF3DEX_GBI_2E -fno-strict-aliasing -fwrapv -DEXT_OPTIONS_MENU -DFS_BASEDIR=""res"" -o build/us_pc/src/game/envfx_snow.o src/game/envfx_snow.c
src/game/mario_cheats.c:839:37: error: 'motos_geo' undeclared here (not in a function); did you mean 'mips_geo'?
{ bhvMotos, motos_geo, "MOTOS", 0, 40, 300, 0x00000000, 1, 0 }, // SPAMBA_MOTOS,
^~~~~~~~~
mips_geo
cc -c -O2 -I include -I build/us_pc -I build/us_pc/include -I src -I . -DDYNOS -DRAPI_GL=1 -DWAPI_SDL2=1 -DAAPI_SDL2=1 -DCAPI_SDL2=1 -DHAVE_SDL2=1 -I/usr/include/SDL2 -I/opt/vc/include -I/opt/vc/include/interface/vcos/pthreads -I/opt/vc/include/interface/vmcs_host/linux -D_REENTRANT -DVERSION_US -D_LANGUAGE_C -DGIT_HASH=""8e2eea10"" -DNON_MATCHING -DAVOID_UB -DF3DEX_GBI_2E -fno-strict-aliasing -fwrapv -DEXT_OPTIONS_MENU -DFS_BASEDIR=""res"" -o build/us_pc/src/game/hud.o src/game/hud.c
src/game/mario_cheats.c:840:37: error: 'friendly_blargg_geo' undeclared here (not in a function); did you mean 'bhv_friendly_blargg_init'?
{ bhvFriendlyBlargg, friendly_blargg_geo, "BLARGG", 0, 40, 300, 0x00000000, 1, 0 }, // SPAMBA_BLARGG,
^~~~~~~~~~~~~~~~~~~
bhv_friendly_blargg_init

[ snipped off more of the same]
`

Help would be appreciated if anyones managed to compile it on the Pi for PC.

Language defaults to Spanish on Linux

Compiled on linux
added both model pack and latest texture pack.

The language is Spanish (I think) and I don't know how to change it.

Game doesn't start with the textures pack

Describe the bug
After compiling and insert the textures in "res" "gfx" the terminal says:

could not determine swap interval, falling back to timer sync
loaded 79 controller mappings from 'db/gamecontrollerdb.txt'
precaching data
unknown texture format: gfx/actors/breakable_box/crazy_box_surface_4xmisc, skipping
unknown texture format: gfx/actors/breakable_box/cork_box_surface.rgba16_renderogmix, skipping
unknown texture format: gfx/actors/breakable_box/cork_box_surface.rgba16_nickelback, skipping
unknown texture format: gfx/actors/breakable_box/crazy_box_surface_nickleback, skipping
unknown texture format: gfx/actors/breakable_box/cork_box_surface.rgba16_16xesrgan, skipping
unknown texture format: gfx/actors/breakable_box/crazy_box_surface.rgba16_renderogmix, skipping
unknown texture format: gfx/textures/sky/rr_textures.02000.rgba16A, skipping
unknown texture format: gfx/levels/ending/cake, skipping
unknown texture format: gfx/levels/ending/cake_luigi, skipping
unknown texture format: gfx/levels/ending/cake_bowser, skipping
unknown texture format: gfx/levels/ending/cake_wario, skipping
could not load texture: gfx/levels/lll/11.rgba16.png
mmap of bo 1242 (offset 0x0000000073837000, size 1048576) failed
Aborted

---

Second start on original build folder the same but few other textures:

could not determine swap interval, falling back to timer sync
loaded 79 controller mappings from 'db/gamecontrollerdb.txt'
precaching data
unknown texture format: gfx/actors/breakable_box/crazy_box_surface_4xmisc, skipping
unknown texture format: gfx/actors/breakable_box/cork_box_surface.rgba16_renderogmix, skipping
unknown texture format: gfx/actors/breakable_box/cork_box_surface.rgba16_nickelback, skipping
unknown texture format: gfx/actors/breakable_box/crazy_box_surface_nickleback, skipping
unknown texture format: gfx/actors/breakable_box/cork_box_surface.rgba16_16xesrgan, skipping
unknown texture format: gfx/actors/breakable_box/crazy_box_surface.rgba16_renderogmix, skipping
unknown texture format: gfx/textures/sky/rr_textures.02000.rgba16A, skipping
unknown texture format: gfx/levels/ending/cake, skipping
unknown texture format: gfx/levels/ending/cake_luigi, skipping
unknown texture format: gfx/levels/ending/cake_bowser, skipping
unknown texture format: gfx/levels/ending/cake_wario, skipping
could not load texture: gfx/levels/lll/5.rgba16.png
could not load texture: gfx/levels/lll/11.rgba16.png
could not load texture: gfx/levels/lll/22.rgba16.png
could not load texture: gfx/levels/lll/12.rgba16.png
could not load texture: gfx/levels/lll/19.rgba16.png
mmap of bo 1250 (offset 0x0000000073939000, size 1048576) failed
Aborted

To Reproduce
Steps to reproduce the behavior:

1. Go to /home/pi/Render96ex/build/us_pc
2. starting "sm64.us.f3dex2e"
3. See error

Expected behavior
A clear and concise description of what you expected to happen.

Screenshots
If applicable, add screenshots to help explain your problem.

Desktop (please complete the following information):

• OS: raspberry pi OS (Bullseye 32bit)

Smartphone (please complete the following information):

• Device: [e.g. iPhone6]
• OS: [e.g. iOS8.1]
• Browser [e.g. stock browser, safari]
• Version [e.g. 22]

Additional context
Add any other context about the problem here.

[Suggestion] Make the RTX feature a switch in the DynOS settings

Describe the suggestion
Can you make it so that i can enable RTX in the DynOS settings

I can't compile it on switch

In the part that I put "source $DEVKITPRO/switchvars.sh" I get "-bash: /opt/devkitpro/switchvars.sh: No such file or directory"

TAS Playback Broken

Describe the bug
When using the TAS playback (cont.m64) it does not play correctly. It seems to do some wrong inputs on file select and then delayed(?) because it doesn't change the camera like the working playback in a repo like sm64ex or sm64rt.

To Reproduce
Steps to reproduce the behavior:
Make sure save file 1 is empty and will play the intro upon selection
Download the old US 1 key TAS http://tasvideos.org/userfiles/info/37135408704300908
Rename to cont.m64 and put in the same directory as sm64
Run Render96

Expected behavior
For the TAS to match console, emu, and other ports

Video
Broken (Render96ex alpha)
https://cdn.discordapp.com/attachments/728112163297951835/844066816498466856/broken.webm

Working (sm64rt)
https://cdn.discordapp.com/attachments/728112163297951835/844066863721349150/Working.webm

Desktop (please complete the following information):
Windows 10 2004
6600k
GTX 1060

Coins sometimes glitch

Describe the bug
Sometimes the coins start to flicker weirdly

To Reproduce
Steps to reproduce the behavior:

1. Later find a glitching coin

Expected behavior
I need it to be smooth as it normally is lol

Screenshots

Super.Mario.64.Render96ex.OpenGL.08153516.2023-03-30.18-46-55.mp4

Desktop (please complete the following information):

• Win11
• Laptop (Victus)
• Microsoft edge
• Mod version: 3.2

Additional context
same thing happens with blue coins and red coins

Outdated build wiki for building for switch

DevkitPro no longer provides the devkitpro-pkgbuild-helpers package, but instead dkp-toolchain-vars. The wiki page says to run this command

(dkp-)pacman -S devkitA64 devkitpro-pkgbuild-helpers libnx switch-tools switch-mesa switch-libdrm_nouveau switch-sdl2

it should instead have

(dkp-)pacman -S devkitA64 dkp-toolchain-vars libnx switch-tools switch-mesa switch-libdrm_nouveau switch-sdl2

Just a small thing, but took me a few minutes to figure out. Would be really frustrating if it was someone's first time.

Render96 test-alpha version crashes.

Describe the bug
Render96 crashes on start-up.

To Reproduce
Steps to reproduce the behavior:

1. Open sm64pcbuilder2
2. Select the render96 test-alpha version
   
3. Click the models tab
   
4. Select this:
   
5. Go to the levels tab and select these:
6. Compile it
7. Run the game
8. Game crashes.

Expected behavior
The game was intended to run normally and not crash.

Screenshots

Desktop (please complete the following information):

• OS: Windows 11
• Browser Chrome
• Version 22H2

Smartphone (please complete the following information):

• Device: Samsung Galaxy A33
• OS: Android 13
• Browser: Stock Browser
• Version: If referring to the One UI version, it's 5.0.

Additional context

BuilderLog_v1.21.3_Render96ex-tester_rt64alpha_Failure_2023-02-11_18.33.06.txt

Fix linking issues on parallel builds

While the Wiki often encourages parallel builds (e.g., with make -j4), tools/Makefile is not currently safe for parallel builds as it doesn't currently account for dependencies between programs.

E.g.:

• tabledesign → audiofile
• skyconv → n64graphics

More often than not, trying to compile tools with make -j4 will fail due to tabledesign finishing to compile before audiofile and then failing to link.

While this would ideally be solved by introducing separate rules per program accounting for dependencies, for now I'm forcing j1 for tools.

This fix was already merged upstream as part of PR sm64pc#512

Not working Rumble

For the rumble function, do I have to declare something during compilation?

I use an Xbox One controller on Win10.
I tried with Render96ex-tester_rt64alpha, Render96ex-alpha, and Render96ex-master (both compilations, DirecX and OpenGL).
With some other repos th controller vibrates normally...

Thanks

tester_rt64alpha - mr_i_eyeball doesn't render properly

Describe the bug
I'm using the Render96_Alpha_v3.0rtx branch of the Render96 ModelPack, along with Render96ex (tester_rt64alpha branch). I noticed that the eyeball model isn't rendering properly, but I'm not sure if it's a ModelPack issue, or a Render96ex issue.

Thank you!

To Reproduce
-Warp to a level with mr_i_eyeball e.g. Big Boo's Haunt, Lethal Lava Land, Hazy_Maze_Cave

Screenshots

Make if statement for include malloc.h work on Mac

Currently I have to manually delete include malloc.h from eloquent.h. The if statement does not work. I am asking to fix the if statement.

How can this be build on an Apple Silicon computer?

Im struggling to build this on my M1 Pro machine - would be nice if this information could be included in the wiki - Thanks!

Compile error

Switch Version crashes after the title screen

Describe the bug
The game crashes after the title screen

Running on v1 Switch
AMS 1.0.0
FW 12.1.0

To Reproduce
Build switch nro and move that one with the res folder into sd:/switch/sm64

Expected behavior
Getting to the fileselect screen and playing the game

Desktop (please complete the following information):

• OS:HOS [Horizon OS]
• Version [12.1.0]

Won't compile on macOS (arm64)

Describe the bug
the build will start, assets are extracted, then it fails here:
n64graphics_ci_dir/exoquant/exoquant.c:28:10: fatal error: 'malloc.h' file not found

To Reproduce
Steps to reproduce the behavior:

1. Clone repo
2. copy baserom
3. run gmake OSX_BUILD=1 BETTERCAMERA=1 EXTERNAL_DATA=1 NODRAWDISTANCE=1

Expected behavior
Should build

Desktop (please complete the following information):
-Macos 13.3.1 (M1 MacBook Pro)

No language files found

Alpha fails to launch with error "No language files found, aborting."

Running windows 7 64 bit

Tried clearing repo, reinstalling dependencies, compiled multiple times with no errors

Many textures don't render on Intel Arc

Describe the bug
At commit 2a66f0f, using the RT64 renderer, Intel Arc, both A750 and A770 16GB, are missing textures. The textures which are either missing or misrendered include all of the fonts except for the "NEW" logo on the main menu, and the Press Start on the title screen. The hand cursor on the file menu is also missing.

Managing to get in game anyway and start a level, the star display screen showing which star I will be getting is glitched as well.

I am using Intel's 31.0.101.4032 drivers.

To Reproduce
Steps to reproduce the behavior:

1. Build the game using sm64builder2 with RT64 renderer and the Render96 model pack.
2. Start the game.
3. Press start on the title screen.
4. Menu is hard to operate without cursor or readable text.

Expected behavior
The textures should work and the menu should be operable.

Screenshots

Desktop (please complete the following information):

• OS: Windows 11 22H2
• Browser: Firefox 108.0.2
• Version: (game version?) 2a66f0f

Additional context
No idea what's causing this, since the presence of XeSS in the game engine would seem to indicate that it's been tested on Arc hardware.

make RENDER_API=RT64 -j4 not working

make RENDER_API=RT64 -j4 not working
ivanb@DESKTOP-3JDV5U1 MSYS ~/render96ex
$ make RENDER_AP
make: cc: No such file or directory
make: *** [Makefile:47: n64graphics] Error 127
Traceback (most recent call last):
File "/home/ivanb/render96ex/./extract_assets.py", line 286, in
main()
File "/home/ivanb/render96ex/./extract_assets.py", line 156, in main
subprocess.check_call(
File "/usr/lib/python3.9/subprocess.py", line 373, in check_call
raise CalledProcessError(retcode, cmd)
subprocess.CalledProcessError: Command '['make', '-s', '-C', 'tools/', 'n64graphics', 'skyconv', 'mio0', 'aifc_decode']' returned non-zero exit status 2.
Makefile:204: *** Failed to extract assets. Stop.

Analogue Camera Not Working

Analogue camera not working on the most recent build using sm64pcbuilder. On windows, right stick just acts as C stick even with analogue camera enabled. Free camera works fine when enabled. Any idea how to fix this?

I can't compile for switch

Hi, I'm trying to compile for the switch with commit 5e0b19e (as indicated by Craftyawesome, here: #30), but when I run the command :

make TARGET_SWITCH=1 HIGHFPS=1 TEXTURE_FIX=1 -j4

this is what I get:

extracting actors/amp/amp_body.rgba16.png
Traceback (most recent call last):
File "./extract_assets.py", line 286, in
main()
File "./extract_assets.py", line 242, in main
subprocess.run(
File "/usr/lib/python3.8/subprocess.py", line 493, in run
with Popen(*popenargs, **kwargs) as process:
File "/usr/lib/python3.8/subprocess.py", line 858, in init
self._execute_child(args, executable, preexec_fn, close_fds,
File "/usr/lib/python3.8/subprocess.py", line 1704, in _execute_child
raise child_exception_type(errno_num, err_msg, err_filename)
OSError: [Errno 8] Exec format error: './tools/n64graphics'
Makefile:208: *** Failed to extract assets. Stop.

Do you have an idea or a solution to this problem ?

The repository is huge to clone (1.5GB)

The git history pack file is around 1.5GB, this should be pruned as it causes unnecessary traffic when cloning the repo

Mario A save file is written as Mario B

I only noticed this once I moved the game folder to my D:\ drive, however I think this did not happen when the game folder was in the original compiler path. This effect does not appear on other save screens.


The Mario A save file I had when I was running the game in the original folder disappeared when I moved the game location, but other options stayed the same.

Edit: I have just learned, the hard way, that using the first Mario B save, it will delete your data when the game is closed. 21~ stars gone.

Language files do not load

Describe the bug
Languages fail to load with an error "Loading File: No language files found, aborting."

To Reproduce
Run the executable under Linux (ie. flatpak bundle)

Expected behavior
No error

Screenshots
n/a

Desktop (please complete the following information):

• OS: SteamOS
• Browser n/a
• Version 3.0

Smartphone (please complete the following information):
n/a

Additional context
The problem is located here:

Render96ex/src/pc/pc_main.c

Line 232 in d0d88b9

alloc_dialog_pool(argv[0], gamedir);

You can't assume that argv[0] always contains the full path. In this situation it only contains the filename.

Game does not compile (Ubuntu 22.10)

Describe the bug
Does not compile

To Reproduce
Steps to reproduce the behavior:
Clone render96 repo and extract the models into the actors folder

Expected behavior
Game to compile successfully without erroring at something about wario's metal cap

Screenshots

Desktop (please complete the following information):

• OS: [Ubuntu]
• Version [22.10]

Putting the r96 texture pack in res folder makes the app not start

Describe the bug
when i put the gfx folder in the res folder it makes the ap not start

To Reproduce
Steps to reproduce the behavior:

1. Go to wiki
2. Click on How to compile eng linux
3. compile as normal
4. open the terminal and run git clone https://github.com/pokeheadroom/RENDER96-HD-TEXTURE-PACK
5. put the gfx file in res
6. try to run the app

Expected behavior
The app opens and the texture pack works

Screenshots
Not applicable

Desktop (please complete the following information):

• OS: [ChromeOS]
• Browser [Chrome]
• Version [90.0.4430.100]

Smartphone (please complete the following information):
don't have one
Additional context

No Wii U Port?

I've been spending days on trying to figure out how to install Render 96 onto my Wii U. I had to go to a discord for 2 days just to get the sm64ex port working aka the one I was told would work with render 96. Then I tried using render96ex only to find in the make file that the is no option for the Wii U. I just don't get how people on YouTube show it on their Wii U, but they NEVER tell you how you can get it. I just don't get it. Please help me with this. I really want render 96 on my Wii U and end my troubles with this.

Switch build crashes after intro

Describe the bug
tried several things like changing the build flags recompiling on linux recompiling on windows cloning a different branch etc.
No matter what I do it's always crashing after the "It's a me Mario" Screen

Running on v1 Switch
AMS 0.19.3
FW 12.0.2

To Reproduce
Build switch nro and move that one with the res folder into sd:/switch/sm64

Expected behavior
Getting to Fileselectscreen

Desktop (please complete the following information):

• OS: HOS (Horizon OS)
• Version [12.0.2]

Render 96 v2.2 DynOs not Saving Model Setting

Describe the bug
The new low poly models are loaded on each new startup even though I set the model pack in DynOS to Render96. To clarify, it does change the models when I change the setting, but it doesn't save when I close the game. If I restart the game, the models reset to their low poly counterparts. Before the latest update, 2.2, there weren't any low poly options included in Render96, so I am wondering if this is just an issue with DynOS loading model packs or if it's specific to Render96.

To Reproduce
Steps to reproduce the behavior:

1. Load game and enter DynOS
2. Open Model Packs
3. Change the Render96 option to "Enabled"
4. Close the game
5. Re-open the game

Expected behavior
I expected the model setting to save between closing and startup as other DynOS settings do (e.g. the Cheater settings save)

Desktop (please complete the following information):

• Windows 10
• Render96ex alpha build v2.2

Note
This is not a major issue, since the models can be changed properly after each launch; it's just a minor inconvenience.

Switch Compile Issue

this is what happens to me. I'm using the correct model pack and commit that works with the switch build, as I learned from another post.

The switch port's crash

Hi everyone, just wanted to know if the nintendo switch version of Render96 got fixed

building on M1 mac

I spent a long time trying to figure out how to build this on my M1 computer, I found a fix, in the makefile there is this section:

# macOS overrides
ifeq ($(HOST_OS),Darwin)
  OSX_BUILD := 1
  # Using MacPorts?
  ifeq ($(shell test -d /opt/local/lib && echo y),y)
    OSX_GCC_VER = $(shell find /opt/local/bin/gcc* | grep -oE '[[:digit:]]+' | sort -n | uniq | tail -1)
    CC := gcc-mp-$(OSX_GCC_VER)
    CXX := g++-mp-$(OSX_GCC_VER)
    CPP := cpp-mp-$(OSX_GCC_VER) -P
    PLATFORM_CFLAGS := -I /opt/local/include
    PLATFORM_LDFLAGS := -L /opt/local/lib
  else
    # Using Homebrew?
    ifeq ($(shell which brew >/dev/null 2>&1 && echo y),y)
      OSX_GCC_VER = $(shell find `brew --prefix`/bin/gcc* | grep -oE '[[:digit:]]+' | sort -n | uniq | tail -1)
      CC := gcc-$(OSX_GCC_VER)
      CXX := g++-$(OSX_GCC_VER)
      CPP := cpp-$(OSX_GCC_VER) -P
      PLATFORM_CFLAGS := -I /usr/local/include
      PLATFORM_LDFLAGS := -L /usr/local/lib
    else
      $(error No suitable macOS toolchain found, have you installed Homebrew?)
    endif
  endif
endif

on M1, the homebrew folder is /opt/homebrew/ so if you change this to:

# macOS overrides
ifeq ($(HOST_OS),Darwin)
  OSX_BUILD := 1
  # Using MacPorts?
  ifeq ($(shell test -d /opt/local/lib && echo y),y)
    OSX_GCC_VER = $(shell find /opt/local/bin/gcc* | grep -oE '[[:digit:]]+' | sort -n | uniq | tail -1)
    CC := gcc-mp-$(OSX_GCC_VER)
    CXX := g++-mp-$(OSX_GCC_VER)
    CPP := cpp-mp-$(OSX_GCC_VER) -P
    PLATFORM_CFLAGS := -I /opt/local/include
    PLATFORM_LDFLAGS := -L /opt/local/lib
  else
    # Using Homebrew?
    ifeq ($(shell which brew >/dev/null 2>&1 && echo y),y)
      OSX_GCC_VER = $(shell find `brew --prefix`/bin/gcc* | grep -oE '[[:digit:]]+' | sort -n | uniq | tail -1)
      CC := gcc-$(OSX_GCC_VER)
      CXX := g++-$(OSX_GCC_VER)
      CPP := cpp-$(OSX_GCC_VER) -P
      PLATFORM_CFLAGS := -I /opt/homebrew/include
      PLATFORM_LDFLAGS := -L /opt/homebrew/lib
    else
      $(error No suitable macOS toolchain found, have you installed Homebrew?)
    endif
  endif
endif

it will now build using gmake

not sure if there is a way to edit the makefile to set the cflags to the brew prefix rather than setting it explicitly