The PlayStation2 is a system designed almost entirely from the ground up for use with CRT TVs. Like any other game console built around analog video output, it is not designed around pixels or resolution, but scanlines and timing. Yes, there is a way to attach a VGA monitor for the official PS2 Linux toolkit so there are some VESA display modes as well, but this is more of an afterthought, and almost no real commercial game used this.

Acronyms used in this article: PS2 (PlayStation2), CRT (Cathode Ray Tube, old TVs that were replaced by HDTVs), GS (Graphics Synthesizer, the PS2’s GPU), PAL (European television signal for CRT TVs), NTSC (American television signal for CRT TVs in America/Japan), CRTC (Cathode Ray Tube Controller, the PS2 GPU has this), EDTV (Enhanced Definition TV or Extended Definition TV, an SDTV that supported progressive scan display modes), VU0/VU1 (Vector Unit 0/Vector Unit 1, two SIMD coprocessors of the PS2)

Sticking locked to 60fps to avoid bad image quality

The PlayStation2’s 4MB embedded VRAM for the Graphics Synthesizer (its GPU) was usually not big enough to hold a full 640×480 framebuffer (or something even bigger). Sony would insist developers not think of it as VRAM but rather a scratchpad, but there’s still only so much you can do with 4MB for render targets.

However, the bandwidth for the GS was unmatched, things like alpha blending, multipasses and framebuffer copies which are expensive on most other GPUs was nearly free on the PS2. In fact, many games like Driv3r abuse some of the GS’ strengths in ways that would bring any other GPU down to its knees. Thanks to its fully programmable geometry pipeline because of the two vector units (VU0 and VU1), the PS2 had hardware features like mesh shaders that we have only started to see being introduced on hardware like the Geforce RTX 20 series almost 18 years later. But we digress.

Developers were almost incentivized by the hardware to make sure their games maintained a solid 60Hz/60fps on NTSC (or 50Hz/50fps on PAL). How did Sony do this? It was perhaps not an intentional initiative but the end result was the same. For one, early versions of the SDK only supported interlaced scanline modes which required 60Hz to get 640×448. Later on, game developers had the option of either using frame mode (full frames) or field rendered mode (interlaced frames).

Field rendering being interlaced means the memory requirements per frame could be halved because the frame is output at 640×240 or even as low as 512×224. This was important because like we said before, the PS2’s GPU only had 4MB of embedded DRAM to work with, and you had to store your framebuffers there. A further benefit is the time needed to render the final output image is also reduced. So to many developers it seemed like field rendering mode was the way to go if you wanted to make a fast performing game on the PS2. So where is the catch? Here comes the big caveat and why PS2 games rarely skip a beat and try to ensure a perfectly locked 60fps. If you miss a frame and the previous one has to be displayed twice, you will see the whole image shift position Y by 1 line. It was therefore imperative to ensure this did NOT happen. So what most games would try to do is instead internally slow down the game by skipping frames when the game was in danger of not maintaining its 60fps target (like SSX 3) but continue to aim at a 60Hz/60fps target.

The other option was frame mode, rendering full frames. Render times obviously go up compared to field rendered mode because you’re rendering full fat frames (640×448 or 512×448), and therefore it might be harder to hit a consistent 60fps vs. field rendering mode. However, the system is more forgiving when you don’t render a new frame in time. The screen would just end up showing the second field from the previous frame.

So, to recap, if the game could maintain a consistently frame paced 60fps for a game, field rendering mode (i.e. interlaced mode) would look as expected, your CRT would blend the half frames and make it look like a full frame. The average enduser would be none the wiser about all the internal dealings of how a CRT assembles and finally outputs a picture, and you have the big advantage that this mode is faster than frame mode and therefore it’s easier to aim for a high framerate like 60fps. If the frame rate would be all over the place, it would lead to a bad picture like discussed in the earlier paragraphs, therefore it was on the developer to make sure you either slow the game internally down or you make sure it runs at a rock solid framerate.

Field rendering being a viable and fast option like this for PS2 as long as you could ensure 60fps/60Hz meant CRT saved the day and PS2 could get away with its generally unimpressive display resolutions.

It has not gone unnoticed by many that PlayStation2 has by far the largest amount of 60fps/60Hz games available (especially on launch day). Turns out there was a technical reason behind this that forced developer’s hand as much as it was developers really wanting to push themselves to the limit.

So if you ever wondered why PS2 games seem to so rarely miss their frametime targets while framerates jump all over the place on other consoles (even Xbox OG and GameCube, its two competitors from that generation), now you know.

We remember at the time PS2 launch games drawing heavy criticism at launch for jagged edges ( colloquially referred to as ‘jaggies’) and the lack of anti-aliasing, especially compared to the Dreamcast. What compounded this problem was that many game magazines and journalists could only do single frame capture. So when they took screenshots or pictures, they would only get half the fields in their pictures. So you’d get bad screenshots in magazines with only the odd or even lines being shown, leading to PS2 games looking much more jaggy in print than they actually did on a real screen. In reality the problem was not so bad, but the lower output resolution to fit in GS eDRAM probably did not help with the misunderstandings and misconceptions.

Widescreen, CRTs and the PS2

Let’s discuss widescreen display modes now and their use on CRT TVs. While a few PlayStation1 games had been adventurous and started implementing widescreen modes, it could be safely assumed that up until now the vast majority of console games were all designed around 4:3 aspect ratios. However, this was soon about to change, thanks in no small part to the PS2. The PS2 doubled as a DVD player and the term ‘anamorphic widescreen’ started being thrown around a lot. Widescreen 16:9 CRT TVs started becoming more mainstream around the early to mid ‘00s.  Most games made for PS2 were still designed around a 4:3 aspect ratio of course, but gradually more and more started offering built-in widescreen modes as demand grew. There are 3 general ways the PS2 can go about displaying a widescreen picture (or in general really):

• Hor+ (Hor Plus) – Extends the horizontal sides of the screen to fill out the screen. Usually the best approach (and almost never used on the real PS2, which is why we need all these widescreen patches)
• Vert- (Vert Minus) – Chops off some of the top and bottom sides of the screen while zooming in on the center
• Hor+ and Vert- (Hor Plus + Vert Minus) – A bit of both

The vast majority of PS2 games unsurprisingly opt for the ‘worst’ of these 3 options when implementing a widescreen mode, Vert-. In some rare cases you will see the 3rd option (Hor+ and Vert-) used as well. Games like Tekken 5 therefore have a ‘quasi-widescreen’ mode, where you lose parts of the screen at the top and bottom that are considered unimportant to the gameplay but then it zooms in slightly in the center to make it fit within a 16:9 aspect ratio. All Ratchet & Clank and Jak and Daxter games are Vert- as well.

It was of course possible to do a Hor+ widescreen game on a PS2, but the developer had to consider system resources, GS embedded VRAM usage, etc. Zooming/scaling was free on the GS and chopping off some parts of the image prob made sure things still fit within the 4MB GS VRAM, so Vert- was probably the easiest solution. With Hor+ also the horizontal resolution comes into play. The more you extend the horizontal sides of the screen, the higher your horizontal resolution needs to be to still maintain a good picture quality. And the PS2 already relied on the CRT’s ability to make interleaved scanlines look really good to get away with lower than average resolutions.

Below you see Tekken 5’s built-in ‘widescreen’ mode. This is how it would look like on a real PS2 with a widescreen CRT TV. As you can see, this shows all the signs of a Vert- aspect ratio. Parts of the top and bottom are cropped while the image is zoomed in on the center. The two characters appear much larger than they would in 4:3 mode. Widescreen on 6th generation consoles was often an exercise in frustration for the enthusiast. So many games would simply give you Vert- and not really take advantage of the extra screen estate.

Below you will see the ‘correct’ widescreen mode, this is an internal widescreen patch in the LRPS2 core. It corrects the widescreen mode to Hor+. Now more of the game world is being rendered onscreen because the image is extended horizontally, no cropping is being done.

Progressive scan

The PS2 released near the tail end of the CRT TV’s life. HD-ready LCD TVs would not start arriving until 2005 and so TV makers were still trying to push the CRT forward in ways they could. As we have established before, CRT TVs are analog video. To make the move to digital television (DTV) on a CRT, new specifications such as EDTV were necessary. EDTV stood for Enhanced-definition television or Extended Definition Television. Basically, in practice it meant that SDTVs were being made that could support 480p and 576p-line signals in progressive scan. Around 2001, progressive scan-capable CRT TVs started being sold, and games started taking advantage of this.

You needed either a component cable (on NTSC TVs) or RGB SCART cables (Japanese and European TVs) in order to be able to use progressive scan display modes. Composite and RF-AV cables did not support this feature.

By pressing and holding X and Triangle at startup, a progressive scan-supported game would ask the user to choose between normal and progressive scan mode. Progressive scan modes are non-interlaced full frame modes, therefore there would be none of the interlacing artefacts and you’d get full-height backbuffers.

The only disadvantage some of these progressive scan-capable games usually have is that sometimes they reduce the framebuffer depth to 16bpp or lower so that everything can still fit inside the Graphics Synthesizer’s 4MB embedded eDRAM. This is not a general rule but some games do make that tradeoff, so you’re trading no interlacing artefacts for a somewhat less high quality final output image (I.e. more color banding).

On average though, progressive scan would still look better than interlaced mode to most people. Some games like Valkyrie Profile 2 and Gran Turismo 4 would even offer 1080i progressive scan modes. Note that this is a bit deceptive. This mode doesn’t actually make the PS2 output at a full 1920×1080 resolution, it’s just a bit of advanced framebuffer shenanigans to make it ‘look’ higher res than it actually is. For these games, 480p progressive scan mode usually looks better on modern displays.

Let’s for instance analyze Gran Turismo 4’s ‘1080i’ mode. The internal render resolution is actually 640 x 540. The GS’ CRTC then magnifies this to ‘1920×1080’. 640 is ‘magnified’ to 1920 via a Magnification Integer (MAGH) of ‘3’. So 640 * 3 ‎ = 1.920. The vertical resolution 540 meanwhile is magnified either by a Magnification Integer (MAGV) of ‘2’ (540 x 2 = 1080), or the interlaced framebuffer switch. So it’s basically a bit of GS CRTC zoom scaling going on. On a CRT this probably looked convincing enough at the time.

Europe and progressive scan (or the lack thereof)

Note that for European versions of the game, progressive scan modes would sometimes be stripped out or removed (God of War 2, Soul Calibur 3). This was probably because of the low adoption rates of progressive scan-capable TVs.

Europe and PAL CRTs (50Hz vs 60Hz)

In fact Europeans had plenty of other issues on top. For those that don’t know, in Europe they used PAL signal CRT TVs while in Japan and the United States, NTSC signal CRT TVs were the standard. PAL ran at 50Hz while NTSC ran at 60Hz.

When PS2 launched in Europe, most Dreamcast games already offered the choice between a PAL 50Hz mode and a PAL60 mode. PAL60 (on TVs that supported it) would give a 60Hz image, thereby avoiding the 16.9% reduction in frame rate for many PAL conversions and the additional letterboxes. The added letterboxing was because PAL typically had a higher output resolution than NTSC. However, games most of the time would not bother taking advantage of this as they were either already nearing the ceiling on system resources or simply because they didn’t care about the European/PAL market enough to cater to them.

The situation on PS2 was more complicated. PAL60 was not a real standard so Sony refused to back it. This meant that most PS2 launch games (prob all of them) sadly did not have any 60Hz options, so we were stuck with 50Hz for a while.

Developers that were usually good about providing PAL-optimized games were UK developers like Psygnosis (Wipeout, Destruction Derby), Core Design (Tomb Raider) and Rockstar/DMA Design (Grand Theft Auto). You’d get a 50Hz mode with more scan lines being rendered than the NTSC version. So you’d get better image quality than the NTSC version. However, the game would still run slower because of 50Hz. Some developers would compensate for this by tweaking the game speed, but regardless in general it would almost usually always be worse than the same game being ran at 60Hz.

No PAL60 for PS2, only NTSC for you

Anyway, around 2002 more and more games started having options at startup allowing you to choose between a 50Hz and 60Hz mode, like ICO. How the PS2 dealt with this compared to the Dreamcast was that instead of allowing you to choose between PAL and PAL60 modes, the game would try to switch to NTSC 480i mode instead. Most televisions sold in Europe near the late ’90s actually supported both PAL and NTSC, so this wasn’t much of a compatibility issue. The games that still didn’t offer these PAL/NTSC selectors (like Silent Hill 2 and Metal Gear Solid 2) would put a bit more effort into their PAL conversions and not have the dreaded letterboxing. The 50Hz/60Hz mode was not an easy thing to crack for many developers. Developers like Square would complain about having to ship a 50Hz and 60Hz version of a FMV scene, and because their games were so front loaded with high-quality big FMV scenes, this was not really possible to fit on the DVD. This is the reason why games like Final Fantasy X sticked to 50Hz despite the developer being aware of the growing demand for NTSC 60Hz mode. Eventually the games that would not have these 50Hz/60Hz toggles would become the exception to the rule.

Moving to LCD/HDTVs in the mid ‘00s for the 7th generation

The industry-wide move from CRTs to LCD TVs was a rough one. This happened sometime around 2005 when new consoles were being prepared. The PS2 of course was still around and wouldn’t really die until like late 2007. There were definitely advantages for the upcoming 7th generation consoles, like the PlayStation3 and Xbox 360. Gone were the days of PAL vs. NTSC. Europeans would no longer have to worry if a game would implement a PAL60/60Hz mode. As long as you hook up your HDMI-capable game console to the HDMI-capable display, your game would output at 60Hz. And of course there was the promise of non-interlaced high resolutions by default. Not a lot of people ever had a progressive scan-capable TV, so this would be their first time seeing a 480p or 720p image on a television.

It would take time for people to catch on, but early LCD HD-ready TVs were often of very poor quality with high amounts of latency and ghosting. CRT-based game consoles (such as the PS2) looked especially poor on them. In fact, for the older sixth generation consoles, there were almost no advantages to be had using them on a non-CRT display. Feedback blur (used to great effect in many PlayStation2 games for motion blur effects) would look great on a CRT, but looked disastrous on these early LCD screens with the heavy ghosting they were known for.  Some games like Soul Calibur 3 would have an in game setting (Software Overdrive) that would try to reduce the afterimage effects on LCD screens.

There was no fixing the latency though or the lack of motion clarity, and this problem would persist for decades. Motion clarity in fact is only now finally being addressed on modern displays for these consoles with the arrival of BlurBusters’ ‘CRT beam racing simulator’ (see our article here for more details). On a modern OLED screen in 202x we can finally enjoy near CRT-like latency with the motion clarity of a CRT (CRT beam simulator shader) AND the looks of a CRT (any advanced CRT shader).

RetroArch 1.19.1 has just been released.

Grab it here.

Remember that this project exists for the benefit of our users, and that we wouldn’t keep doing this were it not for spreading the love to our users. This project exists because of your support and belief in us to keep going doing great things. We have always prioritized the endusers experience, and unlike others, we have never emburdened them with in-app ads, monetization SDKs or paywalled features, and we intend to continue to do so. If you’d like to show your support, consider donating to us. Check here in order to learn more. In addition to being able to support us on Patreon, there is now also the option to sponsor us on Github Sponsors! You can also help us out by buying some of our merch on our Teespring store!

Changelog 1.19.1

• (WASAPI) Only write when running and fix deadlock

We have the pleasure of announcing a great new emulator core, the Neo Geo emulator Geolith. The upstream author of the Jolly Good Emulation project broke out this core and ported it to libretro to make it available to as many users as possible, and we’re happy to welcome it into the fold.

Historically, Neo Geo emulation has fallen under the general ‘arcade’ umbrella and has been handled very capably by FBNeo and MAME. However, arcade emulation is significantly more complex than the console emulation that most users are accustomed to (with ever-changing ROMsets consisting of many tiny files in strangely named zip files to keep up with; specific jargon, like “parent ROMs” and “non-merged sets”; and additional files that sometimes go alongside the game ROMs and sometimes need to live in other places), and this added complexity can be enough of a hurdle that some users struggle to have a good experience with all of the bells and whistles they have come to expect.

This is exactly where Geolith stands out, as it uses static, one-file-equals-one-game ROMsets in the .NEO format developed for the NeoSD, and also used by the FPGA-based MiSTer Neo Geo core (note: Geolith is NOT compatible with ROMs in the MAME/FBNeo zip format). So, if you have dumped your games to be compatible with those devices, they should benefit from full libretro database support and all that entails. That is, simple, one-click scan-and-import into a properly named playlist (with properly named games; you’ll see ‘Art of Fighting’ instead of ‘aof’ without any esoteric fiddling with DATs and so forth), with appropriate icons and boxart/thumbnails automatically applied. This separate, dedicated database and playlist makes it easier to find the games you want to play with less time devoted to setup.

Users who cleared the complexity hurdle of arcade emulation have grown accustomed to FBNeo’s tight and feature-filled libretro integration, which sets a very high bar for Neo Geo emulation by providing for a fast, accurate and bug-free experience with full support for some of RetroArch’s more finicky but high-value features, such as rollback-based netplay and runahead/preemptive-frames. Geolith’s author has put in the effort to provide a comparable, robust experience, so users don’t have to compromise when using this new core.

We’re still polishing up a few rough edges on the database (about 20% of games don’t pull thumbnails yet). Just make sure you stop by the online updater and update your core info files and databases before you get started, since Geolith uses some stuff that was added after RetroArch’s 1.17.0 stable release.

UPDATE: Geolith core is now available on Steam too!

RetroArch 1.17 expands Netplay to support playing multiplayer games that were originally played on two or more separate consoles or computers connected together. This is currently supported by two cores: DOSBox Pure and gpSP.

In DOSBox Pure it is now possible to play DOS and early Windows multiplayer games that used IPX, Ethernet or a modem over the internet.

In gpSP, games that support the Wireless Adapter can also benefit from this feature, although only a subset of games work well at the moment (including Pokemon Fire Red/Green Leaf/Emerald). Those curious about more details can read David G.F.’s article about the gpSP implementation here.

In order to use this new functionality just use the regular Netplay menu, where you can host and/or join netplay games. It works seamless with the existing Netplay feature (including lobbies) with no additional settings required. Make sure however, that you use the same or compatible games, since this won’t be automatically enforced.

RetroArch 1.17.0 has just been released.

Grab it here.

Remember that this project exists for the benefit of our users, and that we wouldn’t keep doing this were it not for spreading the love to our users. This project exists because of your support and belief in us to keep going doing great things. We have always prioritized the endusers experience, and unlike others, we have never emburdened them with in-app ads, monetization SDKs or paywalled features, and we intend to continue to do so. If you’d like to show your support, consider donating to us. Check here in order to learn more. In addition to being able to support us on Patreon, there is now also the option to sponsor us on Github Sponsors! You can also help us out by buying some of our merch on our Teespring store!

Changelog

There are more changes than mentioned here, but for the sake of brevity, here is the changes since 1.16.0.

•  
• ACCESSIBILITY/TTS: fix target language and missing espeak handling on Linux
• AI: AI service reworked: performance increase, automatic translation, configurable subtitle placement, and more
• APPLE: Fix WebDAV crash with digest auth (Cloud Sync)
• APPLE: Cloud sync fixes – ignore .DS_Store files and re-sync on app foreground
• APPLE: Don’t re-create default directories
• APPLE: Show Git information correctly
• AUDIO/SYNC: Handle Hz skew adjustment for high refresh rates better (BFI, swap interval)
• AUDIO/MIXER: Separate ffmpeg/mpv and audiomixer conditionals
• AUDIO/WASAPI: Reworked shared buffer operation for more flexibility, fixed exclusive mode last buffer looping when entering menu
• CONFIG/OVERRIDES: Fix setting savefile_directory in override file
• CONFIG: Force fixed locale for numbers
• CHEEVOS: Upgrade to rcheevos 11.0
• CHEEVOS: Use rc_client for state management
• CHEEVOS: Don’t track disc changes when achievements are disabled
• CHEEVOS: Reinitialize rewind buffer after loading game with achievements
• CHEEVOS: Inform user when server is unreachable
• CHEEVOS: Fix crash on first load of game with achievements with threaded video
• EMSCRIPTEN: Increase chunk size for reads/writes
• EMSCRIPTEN: Modularize the JavaScript and clean up the web build
• EMSCRIPTEN: Prefix core name with libretro_ for exports
• EMSCRIPTEN: replace dashes with underscores in corenames (fix for vitaquake cores)
• EMSCRIPTEN: Use ZipFS for web player asset bundle
• EMSCRIPTEN: Change default audio rate to 44100
• IOS: Fix widget on IOS17
• IOS: Exit instead of crash on some errors
• IOS: Fix ios-vulkan-ppsspp
• IOS: Enable HDR support on IOS
• IOS: Add option to disable respecting silent mode
• IOS: Allow ignoring safe area (fullscreen over notch)
• IOS: Default to not enabling keyboard gamepad
• IOS: Fix XRGB8888 in GL driver
• IOS: Allow changing the app icon
• IOS: Add HAVE_ACCESSIBILITY to iOS builds
• IOS/TVOS: Create GL context as GLES3 to fix some rendering performance problems
• INPUT: Fix ghost input when setting RETROK_UNKNOWN
• INPUT: Default ‘Bind Hold’ to 0 to prevent problems with controllers not resting at null state
• INPUT: Add a setting to allow turbo d-pad directions
• INPUT: Don’t save mouse buttons to autoconfig
• INPUT/MENU: Add option for merging ‘Hotkey Enable’ device types
• INPUT/mFI: disable secondary_joypad to prevent issues with controllers detected both as HID and mFI
• INPUT/UDEV: Change event detection to polling in udev_joypad
• INPUT/WAYLAND: Use unaccelerated pointer motion to prevent mouse dead zone
• LAKKA: Add new menu options for Switch (overclock, CEC, BT ERTM)
• LIBRETRO: Add environment command to get playlist path
• LIBRETRO/NETPACKET: Switch environment call number from 76 to 78 (retire 76 as it was never used by any core). Simplify broadcasts by removing the option to send to all but one client, use an explicit RETRO_NETPACKET_BROADCAST constant instead. Separate explicit flushing and querying of incoming packet into two operations. Enable a core to specify a protocol version string which can get used instead of core version to determine compatibility between two players. Log and notify a separate message when there is a content crc mismatch while using this interface to convey it being less severe
• MENU: Fix menu analog stick navigation (1.16.0.1)
• MENU: Add option to find thumbnail images by playlist rom filename (1.16.0.2)
• MENU: Don’t process menu events while input is being flushed (1.16.0.2)
• MENU: Prevent proceeding to next bind when single binding (1.16.0.3)
• MENU: Fix touchscreen menu false positives in case of overlap
• MENU: Sort playlists ignoring extension (Playstation 2 should come after Playstation)
• MENU: Treat System Volume Information folder as hidden on all platforms
• MENU: Add Belarusian language option
• MENU: Fix savestate thumbnail behavior when navigating back to playlist while content is still running
• MENU: Fix word wrapped widget length
• MENU: Add help text to controller drivers
• MENU: Fix quit on content close option
• MENU: Fix thumbnails in History for content loaded through Load Content
• MENU: Flexible thumbnail matching (ROM name – database name – short name)
• MENU: Remove legacy thumbnail pack downloader
• MENU/GLUI: Icon corrections
• MENU/OZONE: Thumbnail related fixes (missing thumbnail bar, fullscreen thumbnail flashing, sidebar focus)
• MENU/RGUI: Fix text scaling in 16:9
• MENU/XMB: Background images take precedence over color themes, default image opacity set the same as color theme opacity
• MENU/XMB: Fix segmentation fault when background image is missing
• MENU/XMB: Several corrections and cleanups
• MENU/XMB: Layout/thumbnail fixes – Thumbnail layout adjustments, Header title improvements, Handheld layout adjustments
• NETWORKING/NETPLAY: Add support for joining MITM servers from command line
• NETWORKING/NETPACKET: Interface connection flow improvements
• OSX: Fix mouse support for MelonDS DS on OSX
• OSX: Option to create a portable build
• PATCHES: Add support for XDelta-formatted patches.
• PATCHES: Fix patching for cores that support contentless mode
• PS2: Fix for no sound
• PS2: Fix 0-byte logs and add memory stats
• PSP: Fix memory leak in audio driver
• RUNLOOP: Frame Rest, experimental sleep feature aiming to lower CPU usage and temperature when using certain CPU hungry vsync modes
• RPI: Fix videocore + switchres compile failure
• SCAN: Do CRC check on PSP/PSP(PSN) content
• TVOS: Enable overlay support
• VIDEO/BFI: Black Frame Insertion added to DirectX10/11/12. BFI Hz range now covers every 60hz multiple under 1000hz. Variable Strobe length via new ‘Dark Frames’ option, algorithm to auto select ‘decent’ Dark Frames choice.
• VIDEO/GLSL: Add FinalViewportSize support to GLSL
• VIDEO/GLSL: Change rotation type to int to maximize compatibility
• VIDEO: Use video refresh rate instead of core refresh rate for menu frame limiting
• VIDEO: Limit paused video refresh rate
• VIDEO: Enforce swap interval 1 in menu if vsync is on
• WII: Enable Cheevos for Wii builds
• WIIU: Fix config file and core info reading
• WIIU: Fix “Up” and “Left” directional input for both Analog sticks for GC Adapter
• WIIU: Fix 3 USB controllers (NES/SNES/Retrode)
• WIIU: Fix file descriptor leak
• WIN32: Prefer D3D11 and WASAPI as default video and audio drivers
• WIN32: Fix window size when aspect ratio is < 0
• WIN32: Fix screen resolution always tied to the main monitor
• WAYLAND: Enable fallback for screensaver inhibit via D-Bus
• WAYLAND: Introduce wp_fractional_scale_v1 protocol for proper fractional scaling
• WAYLAND: Update wayland-protocols version to 1.31

This was a fun research sideproject. It did not take any development resources away.

Download this on NexusMods here.

This mod does the following:

• Reduce/eliminate texture LOD (Level of Detail)
• Increase draw distances, less pop-in of mappieces and objects
• Untie special effects/particle effects from mappieces/collisions.

This mod takes a conservative and disciplined approach. It does not seek to alter
the game’s look, it does not add additional lighting, it does not remove spot lights or
remove baked shadows. In other words, it tries to look authentic to the original game,
and yet a still noticeable visual upgrade.

REQUIREMENTS

• A copy of Dark Souls Remastered (PC version)

HOW TO INSTALL

1. Extract the contents of the zip archive to the Dark Souls Remastered directory (usually C:\Program Files (x86)\Steam\steamapps\common\DARK SOULS REMASTERED).
2. Start Dark Souls Remastered from Steam, or run DarkSoulsRemastered.exe from the install dir.
3. This will start Dark Souls Remastered with the InfiniDetail mod

MEDIA

See some before/after screenshots here.

Screenshot #1
Screenshot #2
Screenshot #3
Screenshot #4

Full screenshot gallery here.

TODO
* Make more enemies visible all of the time
* Replace more low LOD fake mappieces with full LOD mappieces
* Fix up more rendering oversights from the original game

TOOLS USED
* DSMapStudio

CREDITS
* Fromsoftserve, for his inspirational videos and modding tips.
* DSMapStudio authors for the amazing tool. This mod would not have been possible without it.

FAQ
* Can it be used with DSR Graphics Overhaul mod?
No. DSR Graphics Overhaul by fromsoftserv is a far more ambitious mod that seeks to overhaul the game’s look and modernize it by removing baked shadows and replacing it with dynamic shadows, and also other artistic changes / edits. This mod has a more narrowly defined scope by comparison in that it wants to remain somewhat authentic and at the same time also remove some of the obvious visual rough edges from the original game that probably should have been addressed by the remastered version in the first place.
* How does it affect performance?
This mod can likely be optimized further, consider this a first pass. However, a performance test was done on Steam Deck. It ran at a smooth locked 60fps at the Steam Deck’s native resolution.

WHAT’S NEXT
InfiniDetail mods for Dark Souls 3 and Sekiro were developed alongside this mod, and will appear in the near immediate future.

This was a fun research sideproject. It did not take any development resources away.

Download this on NexusMods here.

This mod does the following:

• Reduce/eliminate texture LOD (Level of Detail)

• Increase draw distances, less pop-in of mappieces and objects

• Draw some mappieces that were inside the level files but never rendered ingame

This mod takes a conservative and disciplined approach. It does not seek to alter
the game’s look, it does not add additional lighting, it does not remove spot lights or
remove baked shadows. In other words, it tries to look authentic to the original game,
and yet a still noticeable visual upgrade.

NOTE: Mappieces that have shadow source/target disabled have not been altered.
There will be a second version of the mod in the future that will enable shadow
source for certain mappieces. Enabling it for all mappieces at once did not turn
out to be a good idea, so it has to be done judiciously.
Altered objects meanwhile do have shadow sources enabled

REQUIREMENTS

• A copy of Dark Souls II: Scholar Of The First Sin (PC version)

This mod will not work with vanilla Dark Souls II.

HOW TO INSTALL

Easy mode

1. Extract the contents of the zip archive to the Dark Souls II: Scholar Of The First Sin directory (usually C:\Program Files (x86)\Steam\steamapps\common\Dark Souls II Scholar of the First Sin\Game).
2. Start Dark Souls II: Scholar Of The First Sin from Steam, or run DarkSoulsII.exe from the DS2 install dir.
   This will start the InfiniDetail mod with ModEngine (supplied in the zip archive)

Hard mode

There is a way to run this mod without ModEngine too.

1. Extract the contents of the zip archive to a temporary directory.
2. Download UXM (https://www.nexusmods.com/sekiro/mods/26)
   Run UXM, point it to the DS2 SOTF exe (usually C:\Program Files (x86)\Steam\steamapps\common\Dark Souls II Scholar of the First Sin\Game\DarkSoulsII.exe). Click on ‘Unpack’ and wait until it’s done. Close UXM.
3. Take the contents of the ‘InfiniDetail’ directory (see Step 1), and copy them over to the DS2 SOTF directory (usually C:\Program Files (x86)\Steam\steamapps\common\Dark Souls II Scholar of the First Sin\Game).
4. Run UXM, point it to the DS2 SOTF exe (usually C:\Program Files (x86)\Steam\steamapps\common\Dark Souls II Scholar of the First Sin\Game\DarkSoulsII.exe). Click on ‘Patch’ and wait until it’s done. Close UXM.
5. You can now start Dark Souls II: Scholar of The First Sin with the mod applied.

TODO

• Stop enemies from animating at half rate from a far distance
• Improve the quality of light shafts
• Improve cascade shadow render distance (max shadowmap resolution in the game seems to be 4096×4096, would need to be able to go beyond this)

TOOLS USED

• ACB Editor

• DSMapStudio

MEDIA

See some before/after screenshots here.

Screenshot #1
Screenshot #2
Screenshot #3
Screenshot #4
Screenshot #5
Screenshot #6
Screenshot #7
Screenshot #8
Screenshot #9
Screenshot #10
Screenshot #11

FAQ

• Can it be used with DS2 Lighting Engine mod?
  Yes, it works fine. First install the DS2 Lighting Engine mod, then walk through this mod’s installation instructions (How To Install).

WHAT’S NEXT

InfiniDetail mods for Dark Souls Remastered, Dark Souls 3 and Sekiro were developed alongside this mod, and will appear in the near immediate future.

RetroArch 1.16.0 has just been released.

Grab it here.

If you’d like to learn more about upcoming releases, please consult our roadmap here.

Remember that this project exists for the benefit of our users, and that we wouldn’t keep doing this were it not for spreading the love to our users. This project exists because of your support and belief in us to keep going doing great things. We have always prioritized the endusers experience, and unlike others, we have never emburdened them with in-app ads, monetization SDKs or paywalled features, and we intend to continue to do so. If you’d like to show your support, consider donating to us. Check here in order to learn more. In addition to being able to support us on Patreon, there is now also the option to sponsor us on Github Sponsors! You can also help us out by buying some of our merch on our Teespring store!

Changelog

There are more changes than mentioned here, but for the sake of brevity, here is the changes since 1.15.0.

1.16.0

• 3DS: Update __system_initArgv
• 3DS: Update SquirrelJME 3DS Core information along with icons.
• ANDROID: Rewrite input handling to better support AINPUT_SOURCE_STYLUS
• ANDROID: Prevent the android quick tap mouse click emulation while pressing a button on the on-screen overlay
• AUDIO: Fast-Forward Audio Resampling
• AUDIO: Reinit audio on audio sync toggle
• AUDIO: Count audio samples in stats when rate control is disabled
• AUDIO/PULSEAUDIO: Support device list for PulseAudio
• AUDIO/WASAPI: WASAPI Frame Delay fix + cleanups
• AUDIO/WASAPI/MIDI: Frame Delay correction
• ARCHIVE/ZIP: Improve ZIP decompression This reduces the amount of memory Retroarch needs to extract a ROM file. It will only need the size of the ROM plus 128KiB to extract the file from the ZIP. Previously it needed as much as twice that amount if the compression ratio was not great. This is useful on memory constrained platforms and has no impact on platforms with plenty of memory. Handles all cases correctly (with and without MMAP, for cores that require fullpath or not, small and big ZIP files).
• APPLE: Add default keyboard overlay directory to platform_darwin
• APPLE: Update to MoltenVK 1.2.4
• APPLE: Don’t include MoltenVK library in Load Core menu list
• APPLE: Enable AltServer JIT for builds installed by Xcode
• APPLE: The default pthread stack size on Apple is too small for detect_ps2_game, so increase it
• CHEEVOS: Upgrade to rcheevos 10.7
• CHEEVOS: Don’t initialize rcheevos memory unless used
• CHEEVOS: Add progress tracker widget
• CHEEVOS: Eliminate leaderboard tracker stutter
• CHEEVOS: Expand leaderboard visibility settings
• CHEEVOS: Collapse trackers with same value definition
• CHEEVOS: prevent frame step when hitting rewind button while paused in hardcore
• CHEEVOS: disallow video_swap_interval and black_frame_insertion in hardcore
• CRT/SWITCHRES: Don’t always force core aspect ratio
• CRT/SWITCHRES: Fix aspect ratio for tate games on a horizontal screen
• CRT/SWITCHRES: Add PAL threshold option for automatic refresh rate switch
• CRT/SWITCHRES/KMS: Add KMS modeswitch
• COMMAND: Make cmd interface more useful for replay information
• CONFIG: Allow all systems to check for backslashes (Windows) as last slash in path. Improves portable core logic
• D3D11: Fix memory corruption in d3d11_gfx_init
• D3D12: Enable blending when drawing the menu
• D3D12: Move fence signaling out of d3d12_gfx_sync
• D3D12/LIBRETRO: Add D3D12 HW_render support
• DOWNLOADS/THUMBNAILS: Prevent directory creation on bogus thumbnail downloads
• EMSCRIPTEN: Fix Emscripten sleep function/macro
• EMSCRIPTEN/RWEBAUDIO: Fix RWebAudioInit race condition
• EMSCRIPTEN/OPENAL: Make openal default audio driver
• FFMPEG: Fix RetroArch fails to restart streaming when video re-inits and instead starts recording
• FRAMESKIP: Use refresh rate instead of core fps for frameskip timing
• INPUT: Combo hold + ‘enable_hotkey’ correction. Fixed issue with having menu toggle hold combo in different button than ‘enable_hotkey’, which caused ‘enable_hotkey’ to also act as menu toggle if held long enough, and simplified and unified duplicate code in start+select holds to a single function.
• INPUT: input_keyboard_event: Don’t check hotkey binds when device is RETRO_DEVICE_POINTER
• INPUT: Add input_keymaps_translate_rk_to_ascii() for correct character input to input_keyboard_event
• INPUT: input_overlay_poll: Delay clearing INPUT_OVERLAY_BLOCKED flag until there is no overlay input (Avoids stray input after osk_toggle)
• INPUT: Send keyboard events for modifiers before other keys (for correct modifier+key input if hitboxes overlap)
• INPUT: Remember currently set keyboard mapping bits during same config read, because otherwise customized keybinds can get cleared out of the bits on the next iteration, causing keyboard events to get passed to the core when they should get blocked.
• INPUT/AUTOCONFIG: Reinit after updating autoconf profiles
• INPUT/LINUX/UDEV: Log mouse devices in info level
• INPUT/LINUX/UDEV: First working version of udev driver with touchscreen support and gestures.
• INPUT/WAYLAND: Add wayland to input driver list
• INPUT/MENU: move port X binds into retropad binds submenu and add appropriate help text and sublabels to discourage people from messing around in there unnecessarily
• INPUT/MENU: Add menu icons to ‘RetroPad Binds’
• INPUT/MENU/OSK: Allow more keyboard actions with menu osk
• INPUT/REMAP: Remap label fix when no autoconf profile active
• INPUT/REMAP: Add ‘Save As’ option for remaps and overrides
• INPUT/OVERLAY: Add overlay parameter to control x/y separation in auto-scale mode
• INPUT/OVERLAY: Revive/rewrite Keyboard Overlay and OSK Toggle. Add keyboard overlay preset, keyboard submenu, and osk_toggle hotkey. Use overlay caching for osk_toggle.
• INPUT/OVERLAY: Overlay Caching. Adds overlay_cache_ptr to keep a disabled overlay in memory when it’s expected to be shown again. Most input_overlay_deinit calls are replaced with input_overlay_unload, which caches the overlay unless initing/deiniting core or disabling overlays. Loading a cached overlay is done as a swap, intended for osk_toggle.
• INPUT/OVERLAY: Add input_overlay_check_mouse_cursor() to preserve show/hide mouse behavior
• INPUT/OVERLAY: Don’t apply input_overlay_show_mouse_cursor in windowed mode (controlled by mouse grab only)
• INPUT/REMAPPING: Stop always reseting to defaults on remap delete
• INPUT/SDL: Fix rumble on PS4/PS5 controllers connected via bluetooth
• IOS: JIT support in iOS
• IOS: Use AltKit to try to enable JIT on iOS
• IOS: Fix thumbnails in history on iOS
• IOS: Enable option to rumble iOS device instead of controller
• IOS: Fix including audio/video filters on iOS
• IOS: Haptic feedback for overlay on iOS
• IOS: On iOS, stop/start audio on interruptions
• IOS: Add accelerometer and gyroscope sensors to iOS
• IOS: iOS needs to targets 13 due to some keyboard emulation, but tvOS does not
• IOS/TVOS: Rework JIT availability checks
• IOS/TVOS: When updating playlist with detected core path, used resolved core path
• IOS/TVOS: Implement memory queries on iOS/tvOS
• IOS/TVOS: Netplay discovery through bonjour/mdns
• IOS/TVOS: Improve some log formatting on iOS/tvOS
• IOS/TVOS: Preliminary MetricKit support on iOS/tvOS
• TVOS: Fix analog stick handling after adding Siri remote tap support
• TVOS: Also handle tap events on tvOS
• TVOS: Back up retroarch.cfg to NSUserDefaults on tvOS.
• TVOS: tvOS has its own beautiful screensavers and I would like them
• TVOS: tvOS Siri remote handling
• TVOS: Fix accidental left/right keypresses on tvOS
• TVOS/CONFIG: Avoid a crash if there is no backed up config file on tvOS It’s not really usable as a game controller, but it is good for going through the menu as a simple LRUD. This also adds better support for names of mFI controllers, as well as being able to do the expected tvOS behavior of “backing out” of the app.
• TVOS: AltKit support for tvOS
• INTL: Fix language detection in Canada to English
• LANGEXTRA: Enable language autodetect for all builds with LANGEXTRA
• LIBRETRO: Add new context hardware render enums – enables autodetection of DX12 for PS2 core
• LIBRETRO: Add API to check JIT availability on iOS
• LIBRETRO: Allow RETRO_ENVIRONMENT_SET_MEMORY_MAPS also after core startup. Change the comment in libretro.h about the removed limit and handle the environment call during core runtime in RetroArch.
• LIBRETRO/MICROPHONE: Add new API for microphone support.
• LIBRETRO: Add new API for querying the device’s power state.
• LIBRETRO/VFS: Rewrite retro_vfs_file_remove_impl
• LINUX: Input driver fix 8+ joypads. It was reported that controllers beyond 8 worked only partially (analogs yes, but not buttons), and the found fix was also confirmed.
• MIDI: Show MIDI output first
• MENU: Reorganize ‘Saving’ menu
• MENU: Start directory browsing from current value
• MENU: Fix menu toggle combo hold with same ‘enable_hotkey’
• MENU: Add menu scroll home+end actions
• MENU: Move ‘systemfiles_in_content_dir’ from Saving to Core
• MENU: Menu navigation acceleration adjustments
• MENU: Audio synchronization menu cleanup
• MENU: Menu value label cleanup
• MENU: Show playlist history/favorites icons by default
• MENU: Remove advanced option flag from video rotation + orientation
• MENU: Combine audio resampler menu with audio output menu
• MENU: Microphone menu cleanup
• MENU: Add missing menu visibility option for content dir override
• MENU: Fix menu refresh in action_ok_network
• MENU: Video settings menu cleanup
• MENU: Try to declutter Quick Menu
• MENU: Change network port menu options to ‘allow_input’ mode
• MENU: Preselect current core in “Set Core Assocation” menu for more carefree usage
• MENU: Focus back to “Set Core Association” item after returning from the menu instead of first item
• MENU: Replaced hack fix for focusing the first item after returning from core suggestion “Run” menu after association reset
• MENU: Fixed setting core association in history/favorites after resetting it and running without setting
• MENU: Relocate playlist manager core association options to prevent accidental resets with Start
• MENU: Fraction setting wraparound rounding correction
• MENU: Add more missing sublabels
• MENU: ‘Updater Settings’ relocation
• MENU: Search box usability improvements
• MENU/RGUI: Implement ‘Remember Selection’ option
• MENU/RGUI: Fix savestate thumbnail fullscreen cancel action
• MENU/RGUI: Fix and add toggle for playlist thumbnails
• MENU/OZONE: Update Dracula theme
• MENU/OZONE: Return to sidebar from playlist manage quick access
• MENU/OZONE: Sidebar multiline scrolling
• MENU/OZONE: Avoid crash if ozone sidebar_index_size is 0
• MENU/OZONE: Prevent metadata and footer overlap
• MENU/OZONE: Sublabel width tweak
• MENU/OZONE: Mouse hover selection fix
• MENU/XMB: Message/help box adjustments
• MENU/XMB: XMB menu playlist index bugfix
• MENU/XMB: Playlist label limit fixes
• MENU/XMB: Prevent playlist label truncating long items even when right thumbnail is not enabled/visible
• MENU/XMB: Prevent playlist label length resetting to full width on every item change while thumbnail is pending
• MENU/XMB: Stop showing playlist entry index outside of playlists
• MENU/XMB: Fix menu focus issue when returning from “Set Core Association” if playlist index is greater than the previous menu
• MENU/EXPLORE: Prevent flashing when browsing fullscreen thumbnails
• MENU/EXPLORE: Match label ticker length behavior with playlists
• MENU/SOUNDS: Fix certain audio drivers from hanging when menu pause is enabled with menu sounds
• MENU/SOUNDS: Fix menu sounds stopping after fullscreen toggle / video reinit
• MENU/QT/WIMP: Cleanup Desktop Menu welcome text
• MENU/OVERLAYS: Reorder overlay menu
• MICROPHONE: Add support for microphones.
• MICROPHONE/ALSA: Add alsa and alsathread microphone drivers.
• MICROPHONE/SDL: Add sdl2 microphone driver.
• MICROPHONE/WASAPI: Add wasapi microphone driver.
• MOBILE: On mobile in portrait mode, don’t override custom viewport
• NETWORKING: Enhance netpacket interface
• NETWORKING/NETPLAY: Enable core host to refuse connecting new players to limit the number of connected players
• NETWORKING/NETPLAY: Enable a core to flush outgoing packets and read incoming packets without waiting for the next frame (can be used for lower latency or blocking reads)
• OPENDINGUX/RG350: Enable networking for RG350
• OPENGL/WAYLAND: Add frame callback after egl_swap_buffers to improve latency when max_swapchain <= 2
• OSD/STATISTICS: Show current video + audio drivers in statistics
• OSD/STATISTICS: Ensure statistics text is aligned left
• OSX/MACOS: Fix steam OSX core path
• OSX/MACOS: Add HAVE_STDIN_CMD support
• OSX/MACOS: fix mouse grab in windowed mode
• OSX/MACOS/IOS/TVOS: Enable SSL in macos/ios/tvos builds
• PLAYLIST: Change default playlist last played time format
• PS3/PSL1GHT: Add overlay support
• RUNLOOP: Don’t count frames while paused
• SERENITYOS: Add SerenityOS to the list of supported operating systems
• SCANNER: Improve scanner console output
• SCANNER: More scanner console output coloring
• SCANNER: Improve CLI scan console output
• SCANNER: Add content import/scan via CLI
• SCANNER/PS1: Serial extraction is now considered a failure if PSX.EXE is found, allowing CRC fallback and actually fixing the game scanning
• VIDEO: Correct rotated core provided aspect ratio
• VIDEO: Minor adjustments to refresh rate switch behavior
• VIDEO: Fix windowed viewport with libretro rotation
• VIDEO/SDL2: Use “nearest” scaling in menus
• TASKS: Fix corrupt task progress percentage
• XVIDEO/MENU: Add menu support to xvideo gfx driver, making it usable
• UWP: Fix modifier keys ( shift, ctrl, alt ) as well as F10 not being detected on the UWP platform
• VULKAN: Ignore Fast-Forward Frameskip option
• VULKAN/KHR_DISPLAY: Support screen refresh rate with Vulkan KHR_Display context
• WIN32: Increase maximum window limit
• WIN32/WINDOWSXP/MSVC2010: Fix invisible menu display
• WAYLAND: Remove splash screen
• WAYLAND: Check for pointer before locking it
• WAYLAND: Add mouse grab/lock functionality
• WAYLAND: Get system wayland-protocols path via pkg-config

The cause of the intrusion has been found and isolated. We have no indications that other systems including core distribution was accessed, so there is no immediate cause for concern there. Sites should be operational right now and everything should work fine right now.

PSA: If people downloaded RetroArch on Windows earlier today they should check their history for mediafire. Trusted downloads from Libretro always originate from http://buildbot.libretro.com and/or http://bot.libretro.com. We will never upload files to any third party file sharing site.

More to share on this later on.

DirkSimple core added to Steam

We added a new core, DirkSimple. This brings the total number of available cores on Steam to 60 now.

DirkSimple is a core written from scratch by icculus, one of the original creators/maintainers of popular open source projects like SDL. It can play some of the 1980s LaserDisc arcade games.

See our Steam article here.

RetroArch 1.15.0 updated

A maintenance update/fix has been pushed for 1.15.0. This replaces the current 1.15.0 version.

New changes:

• INPUT/AUTOCONFIG: Check for ‘enable_hotkey’ also from autoconf binds
• LIBRETRO-COMMON/VFS/FILESTREAM: Fixes filestream_vscanf regression
• MENU/XMB: Horizontal icon animation fix
• OSD/STATISTICS: Notification font + statistics adjustments
• OSD/STATISTICS: Finetuned statistics layout to be more compact and aligned
• OSD/STATISTICS: Group Run-Ahead and Frame Delay as “Latency”
• OSD/STATISTICS: Try to scale font as small as possible/readable if stats won’t fit
• OSD/STATISTICS/FONT: Allow reseting notification font with RetroPad Y to “null”, which uses the fallback pixel font
• OSD/STATISTICS/FONT: Show “Default” instead of empty with default font
• OSD/STATISTICS/FONT: Start browsing font from assets instead of root
• WIN32: Ignore window limiting with fixed position. The other resizing part already took this into account, but WM_GETMINMAXINFO did not.