fix(vk): the 2x hint names AMD FSR 3.1 only where FSR 3.1 makes the frame, not while DLSS FG makes it #1497

Merged
llun merged 1 commit from pmwork-personal/Longshot:fix/framegen-2x-hint-dlss into main 2026-11-18 01:38:07 +00:00
Member

Type: fix · Risk: low · Tested on: Windows 11, RTX 5090 (NVIDIA 621.24) · Needs: none

TL;DR

With NVIDIA DLSS Frame Generation chosen, the Frame generation row's hint at 2x said the frame was made "by AMD FSR 3.1", while DLSS FG made it. Since #1486 every RTX 40 with DLSS FG runs at 2x, so every such player saw the wrong hint. Now the 2x hint names FSR 3.1 only where FSR 3.1 makes the frame. A new case in the check fails on main and passes here, and the menu showed the right hint in the game.

Breakdown

  • Who it helps: players with DLSS Frame Generation at 2x, above all every RTX 40 (its DLSS FG stops at 2x since #1486).
  • What was wrong: the hint asked only whether FSR 3.1 frame interpolation is wanted, which is also true with DLSS FG chosen (FSR 3.1 is asked then for the sets DLSS FG does not make).
  • What changes: the 2x hint names FSR 3.1 only when DLSS FG does not make the frames.
  • What does not change: which generator makes the frames; every other hint; the help box's height (the FSR text is still measured).
  • How it was checked: the real hint code compiled and run in the check (it fails on main); the full suite against main; the menu in the game on an RTX 5090.
  • What is still open: nothing.
  • Risk: low: one condition in a menu text.

Why

The Frame generation row's hint at 2x read "One made frame between each two of the game's, by AMD FSR 3.1 from the scene's depth and motion." whenever fsr3fi::frames_wanted() said yes. That is yes with Frame generation by at DLSS FG too, because FSR 3.1 is then asked for the sets DLSS-G does not make. So with DLSS FG making every frame, the hint named the wrong engine. Since #1486 a DLSS FG multiplier stops at the runtime's most + 1, which is 2x on an RTX 40, so every RTX 40 with DLSS FG shows this hint.

How to reproduce

Seen on: Windows 11, RTX 5090, NVIDIA 621.24 (any DLSS FG card).
Settings and switches: Upscaler NVIDIA DLSS or DLAA, Frame generation by DLSS FG, Frame generation 2x.
Steps: open the Insert menu and select the Frame generation row.
Expected: "One made frame between each two of the game's."
Actual: "One made frame between each two of the game's, by AMD FSR 3.1 from the scene's depth and motion."

What changed

hn_framegen names FSR 3.1 at 2x only while DLSS-G does not make the frames.

  • menu: hn_framegen (gpu/vk/vk_menu.cpp) shows the FSR text at 2x only when dlss_fg_most() is 0 (FSR 3.1 FG chosen, DLSS FG that does not run here, no temporal Upscaler, or given up) and fsr3fi::frames_wanted() says yes. Its variants for the help box's height still include the FSR text.
  • checks: tests/test_framegen_dlss_cap.py compiles and runs the real hn_framegen with an RTX 40's and an RTX 50's DLSS FG, FSR 3.1 FG, the built-in one, and the measured variants.

Measurements

Verdict: nothing to measure: a menu text, checked in the game and by the check.

Flags and settings

None: one hint text of the Frame generation row changes; no row, ini key or switch.

Testing

Tested: In the game on Windows 11, RTX 5090 (NVIDIA 621.24), 5120x2160, DLSS (DLAA), Frame generation by DLSS FG, Frame generation 2x, a dev pack of this branch on a shared base (scripts/dev_packs.py): the Frame generation row at 2x shows "One made frame between each two of the game's." with no FSR 3.1. Before (the same code as main, from #1486's run): "... by AMD FSR 3.1 from the scene's depth and motion." Game-free: tests/test_framegen_dlss_cap.py fails on main (the DLSS FG cases) and passes here; python -I -B tests/run_all.py fails only checks that fail on main on this PC (53 checks, all failing on main 310e7c9d here too: compilers, headers and stand-ins this PC lacks).

Before: main's code @ 6556a50d (#1486, merged as ec76f696; 2x, DLSS FG chosen, the start screen)

Before: the 2x hint names AMD FSR 3.1 with DLSS FG

After: this PR @ 8469aedd (the condition's order changed since, not its result) (2x, DLSS FG chosen, the start screen)

After: the 2x hint without FSR 3.1

Not tested: an RTX 40 card (this PC has an RTX 5090; the condition is the same on both, and the check runs an RTX 40's case); Linux.

Must be tested:

  • None.

Risks

What could break: nothing outside the hint's text.

Turning it off: revert the commit.

Compatibility: nothing changes.

Checklist

Everyone

  • One topic.
  • The title is a Conventional Commit (type(scope): description) of the status line's type.
  • Why gives the reason plainly and links the issues that fit (the tracker searched).
  • Tested on my own PC where it runs (in the game, for a change the game runs), as Testing's Tested says; what
    can't be run by hand is in Testing's Not testable by hand, with why.
  • Measurements has a verdict, or one line saying why there is nothing to measure.
  • Nothing gets worse that I know of: any cost (speed, latency, memory, a setup that worked before) is measured and
    in Risks.
  • Testing names what was tested, what wasn't, and what must still be tested, where and how.
  • Rebased on current main.
  • python -I -B tests/run_all.py passes (any failure that also fails on main is named).
  • No game files, dumps, build output or generated files.
  • third_party/ untouched (a new folder with a PROVENANCE.md instead).
  • The license is OK: GPL-3.0-or-later, nothing incompatible and nothing decompiled.

When it applies

  • AI assistance recorded with Assisted-by: / Co-Authored-By: trailers.
  • Screenshots of the same spot, captioned like the Measurements columns, for a change that shows on screen.
  • A game-free test added for the fix, failing on main and passing here.
  • A check under tests/ that reads changed source updated in this PR, not weakened.

🤖 Generated with Claude Code

**Type:** fix · **Risk:** low · **Tested on:** Windows 11, RTX 5090 (NVIDIA 621.24) · **Needs:** none ## TL;DR With NVIDIA DLSS Frame Generation chosen, the Frame generation row's hint at 2x said the frame was made "by AMD FSR 3.1", while DLSS FG made it. Since #1486 every RTX 40 with DLSS FG runs at 2x, so every such player saw the wrong hint. Now the 2x hint names FSR 3.1 only where FSR 3.1 makes the frame. A new case in the check fails on main and passes here, and the menu showed the right hint in the game. ## Breakdown - **Who it helps:** players with DLSS Frame Generation at 2x, above all every RTX 40 (its DLSS FG stops at 2x since #1486). - **What was wrong:** the hint asked only whether FSR 3.1 frame interpolation is wanted, which is also true with DLSS FG chosen (FSR 3.1 is asked then for the sets DLSS FG does not make). - **What changes:** the 2x hint names FSR 3.1 only when DLSS FG does not make the frames. - **What does not change:** which generator makes the frames; every other hint; the help box's height (the FSR text is still measured). - **How it was checked:** the real hint code compiled and run in the check (it fails on main); the full suite against main; the menu in the game on an RTX 5090. - **What is still open:** nothing. - **Risk:** low: one condition in a menu text. ## Why The Frame generation row's hint at 2x read "One made frame between each two of the game's, by AMD FSR 3.1 from the scene's depth and motion." whenever `fsr3fi::frames_wanted()` said yes. That is yes with Frame generation by at DLSS FG too, because FSR 3.1 is then asked for the sets DLSS-G does not make. So with DLSS FG making every frame, the hint named the wrong engine. Since #1486 a DLSS FG multiplier stops at the runtime's most + 1, which is 2x on an RTX 40, so every RTX 40 with DLSS FG shows this hint. ## How to reproduce **Seen on:** Windows 11, RTX 5090, NVIDIA 621.24 (any DLSS FG card). **Settings and switches:** Upscaler NVIDIA DLSS or DLAA, Frame generation by DLSS FG, Frame generation 2x. **Steps:** open the Insert menu and select the Frame generation row. **Expected:** "One made frame between each two of the game's." **Actual:** "One made frame between each two of the game's, by AMD FSR 3.1 from the scene's depth and motion." ## What changed `hn_framegen` names FSR 3.1 at 2x only while DLSS-G does not make the frames. - menu: `hn_framegen` (`gpu/vk/vk_menu.cpp`) shows the FSR text at 2x only when `dlss_fg_most()` is 0 (FSR 3.1 FG chosen, DLSS FG that does not run here, no temporal Upscaler, or given up) and `fsr3fi::frames_wanted()` says yes. Its variants for the help box's height still include the FSR text. - checks: `tests/test_framegen_dlss_cap.py` compiles and runs the real `hn_framegen` with an RTX 40's and an RTX 50's DLSS FG, FSR 3.1 FG, the built-in one, and the measured variants. ## Measurements **Verdict:** nothing to measure: a menu text, checked in the game and by the check. ### Flags and settings None: one hint text of the Frame generation row changes; no row, ini key or switch. ## Testing **Tested:** In the game on Windows 11, RTX 5090 (NVIDIA 621.24), 5120x2160, DLSS (DLAA), Frame generation by DLSS FG, Frame generation 2x, a dev pack of this branch on a shared base (`scripts/dev_packs.py`): the Frame generation row at 2x shows "One made frame between each two of the game's." with no FSR 3.1. Before (the same code as main, from #1486's run): "... by AMD FSR 3.1 from the scene's depth and motion." Game-free: `tests/test_framegen_dlss_cap.py` fails on main (the DLSS FG cases) and passes here; `python -I -B tests/run_all.py` fails only checks that fail on main on this PC (53 checks, all failing on main 310e7c9d here too: compilers, headers and stand-ins this PC lacks). **Before: main's code @ 6556a50d** (#1486, merged as ec76f696; 2x, DLSS FG chosen, the start screen) ![Before: the 2x hint names AMD FSR 3.1 with DLSS FG](https://git.magicbean.dev/attachments/67bb6b33-ee08-4038-8bef-d6b33d0010c7) **After: this PR @ 8469aedd** (the condition's order changed since, not its result) (2x, DLSS FG chosen, the start screen) ![After: the 2x hint without FSR 3.1](https://git.magicbean.dev/attachments/9204bb1d-631c-4322-879f-ad93a4af17d6) **Not tested:** an RTX 40 card (this PC has an RTX 5090; the condition is the same on both, and the check runs an RTX 40's case); Linux. **Must be tested:** - None. ## Risks **What could break:** nothing outside the hint's text. **Turning it off:** revert the commit. **Compatibility:** nothing changes. ## Checklist ### Everyone - [x] One topic. - [x] The title is a Conventional Commit (`type(scope): description`) of the status line's type. - [x] Why gives the reason plainly and links the issues that fit (the tracker searched). - [x] Tested on my own PC where it runs (in the game, for a change the game runs), as Testing's Tested says; what can't be run by hand is in Testing's Not testable by hand, with why. - [x] Measurements has a verdict, or one line saying why there is nothing to measure. - [x] Nothing gets worse that I know of: any cost (speed, latency, memory, a setup that worked before) is measured and in Risks. - [x] Testing names what was tested, what wasn't, and what must still be tested, where and how. - [x] Rebased on current `main`. - [x] `python -I -B tests/run_all.py` passes (any failure that also fails on `main` is named). - [x] No game files, dumps, build output or generated files. - [x] `third_party/` untouched (a new folder with a `PROVENANCE.md` instead). - [x] The license is OK: GPL-3.0-or-later, nothing incompatible and nothing decompiled. ### When it applies - [x] AI assistance recorded with `Assisted-by:` / `Co-Authored-By:` trailers. - [x] Screenshots of the same spot, captioned like the Measurements columns, for a change that shows on screen. - [x] A game-free test added for the fix, failing on `main` and passing here. - [x] A check under `tests/` that reads changed source updated in this PR, not weakened. 🤖 Generated with [Claude Code](https://claude.com/claude-code)
fix(vk): the 2x hint names AMD FSR 3.1 only where FSR 3.1 makes the frame, not while DLSS FG makes it
Some checks failed
CI / Python source checks (ubuntu-24.04, part 1/2) (pull_request) Has been skipped
CI / Python source checks (ubuntu-24.04, part 2/2) (pull_request) Has been skipped
CI / SPIR-V fixtures (global narrow load/store, direct surface atomics, F2I immediate) (pull_request) Has been skipped
Third-party parts / The same as upstream (pull_request) Has been skipped
CI / Conventional Commit title (pull_request) Successful in 7s
Website / Website on Cloudflare (pull_request_target) Has been skipped
Guide screenshots / Guide images and local UI source review (pull_request) Failing after 10s
CI / Workflow lint (actionlint) (pull_request) Successful in 10s
Third-party parts / As their notes say (pull_request) Successful in 12s
PR description / Description follows the template (pull_request_target) Successful in 4s
d7f20f4fc5
The Frame generation row's hint at 2x said "One made frame between each two of the game's, by AMD FSR 3.1 from the
scene's depth and motion." whenever fsr3fi::frames_wanted() said yes. That is also yes with DLSS FG chosen: FSR 3.1 is
asked then for the sets DLSS-G does not make. Since #1486 an RTX 40 with DLSS FG at a fixed multiplier, on a temporal Upscaler, runs at 2x (its runtime's
most + 1), so every such card showed that hint while DLSS FG made the frame (seen on an RTX 5090 at 2x, too). hn_framegen now
names FSR 3.1 only where DLSS-G does not make the frames (vk_dlss.cpp dlss_fg_most is 0: FSR 3.1 FG chosen, or DLSS FG
that does not run here); the help box still measures that text at 2x.

Check: tests/test_framegen_dlss_cap.py compiles and runs the real hn_framegen (an RTX 40's and an RTX 50's DLSS FG, FSR
3.1 FG, the built-in one, the measured variants); it fails on main.

Co-Authored-By: Claude Opus 5.5 <[email protected]>
Author
Member

Tested in game: RTX 5090, NVIDIA 621.24, Windows 11, 5120x2160, DLSS (DLAA), Frame generation by DLSS FG, Frame generation 2x, the start screen: the Frame generation row's hint shows 'One made frame between each two of the game's.' with no 'by AMD FSR 3.1' (on main it named FSR 3.1 while DLSS FG made the frame).

Tested in game: RTX 5090, NVIDIA 621.24, Windows 11, 5120x2160, DLSS (DLAA), Frame generation by DLSS FG, Frame generation 2x, the start screen: the Frame generation row's hint shows 'One made frame between each two of the game's.' with no 'by AMD FSR 3.1' (on main it named FSR 3.1 while DLSS FG made the frame).
Author
Member

Added the Before/After screenshots of the Frame generation row at 2x with DLSS FG (RTX 5090).

Added the Before/After screenshots of the Frame generation row at 2x with DLSS FG (RTX 5090).
llun approved these changes 2026-11-18 01:38:00 +00:00
llun merged commit a6c8ed5707 into main 2026-11-18 01:38:07 +00:00
Sign in to join this conversation.
No reviewers
No milestone
No project
No assignees
2 participants
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set

Reference
MagicBeanDev/Longshot!1497
No description provided.