Commit Graph
8 Commits
Author SHA1 Message Date
Valiantsin DzerakhandClaude Opus 5.5 22794ab332 Merge task/fg into task/fsr31: frame generation's per-object motion vectors and its contexts kept across dynamic resolution, on one motion gate with FSR 3.1
vk_objmv_wanted is FSR 3.1's latched upscaler half (DLSS with NGX up, or FSR 3.1 picked at start) or frame generation following the scene with its objects (vk_fg_objects_on), and vk_mvprev_frame turns the tables on for the running upscaler or for frame generation. vk_usnap.c keeps this branch's host regions: a region's host flag says whether this submission wrote its host copy, which covers frame generation going on and off live without fg's separate U.host pointer. The jitter fix in mv_pass stays a single copy.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-20 20:15:40 +03:00
Valiantsin DzerakhandClaude Opus 5.5 5e3475c583 With the per-object motion on, the uniform snapshot copies into host memory only the regions the CPU reads back (those holding a vertex-stage block of a bank the motion compares, or of the camera, c4) and writes the rest straight into the device copy, as with the motion off; before, the whole snapshot went through host memory, every byte copied twice. A host region is still copied into host memory first and from there into the device copy, so both hold the same bytes while the game writes guest memory. vk_usnap_bytes reads a host region's host copy and any other from the device copy (slow through the BAR, which only rare readers reach: the merged uniform updates and the test surveys), so every reader gets the bytes it got before. The last-frame copy (vk_usnap_prev) takes host regions only. The snapshot summary gives the bytes copied into host memory.
Field still view, 28 s, TOTK_MVPREV_CHECK=1: mv_check 0 of 60 checks with over 0.5 % object pixels fit better by a variant; mv_prev intact 100 % in c5, c6, c7, c14, c15 and c16; 0 of 214238 last-frame copies differ from the neighbour copies.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-20 19:53:34 +03:00
Valiantsin DzerakhandClaude Opus 5.5 b8edbdd873 The per-object motion reads last frame's object and bone blocks from last frame's uniform snapshot, kept one more frame, instead of having every snapshot copy both neighbours of each such block: the town's snapshot was 2836 KiB a submission with the neighbours and 1916 KiB without, and each byte is copied twice with the motion on (into host memory for the CPU's reads, then into the device copy). vk_mvprev_find still picks the address; the block's bytes and the motion draw's descriptor come from vk_usnap_prev. A block bound at the same address both frames keeps this frame's copy, as before.
Lifetime: a scene submission's arena (vk_usnap_scene, from the scene draws vk_mvprev notes) is kept at its vk_usnap_end as the last-frame copy instead of going to the free list at its fence. The next scene submission's end, or any end once the motion is off, hands the kept arena to the completion thread with that submission's fence, then keeps its own. The motion draws that bind the kept arena are listed and recorded within the next scene submission (vk_mvobj_execute drops a list of another fence), so the GPU has read it before that fence; the fences of one queue complete in order, so its own submission's reads are done too. One arena is held beyond those in flight, whatever the session length; vk_usnap_shutdown frees it with the others. A frame whose scene spans two submissions finds only the last one's blocks; the others keep the camera's motion.

TOTK_MVPREV_CHECK=1 still copies the neighbours and compares them with the kept copy. Field still view, 28 s: the mv_prev lines intact 100 % in c5, c6, c7, c14, c15 and c16; over whole blocks 87 of 218320 last-frame copies differ from the neighbour copy (the game changed the other copy after last frame's submit; the kept copy is what last frame drew); mv_check 0 of 60 checks with over 0.5 % object pixels fit better by a variant, object end-point error median 0.98 px against the camera's 1.27; the c7 lookups split as before (nearer to side 1.87, reference 1.83).

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-20 19:49:47 +03:00
Valiantsin DzerakhandClaude Opus 5.5 82a1f526a9 The per-object motion keeps last frame's blocks only for the banks its listing looks up (c5, c6, c7, c14, c15 and c16 in the town and the field) instead of every bank a variant reads (c5 to c16 but c12): the scene draws stop probing the c8, c9, c10, c11 and c13 blocks into the table and the snapshot stops copying their neighbours, none of which a lookup ever read (c9 and c11 stay at this frame's block, the others belong to shaders that are not drawn again). The table entries of the banks looked up are the same as before, so the vectors are too. The listing's per-shader memo now comes before the shader's resource list and also keeps the filters' verdict, so a scene draw that is no candidate costs one memo probe.
Noting only the candidates' blocks was tried and dropped: the table then misses the other draws' blocks, a sixth of the c7 "nearer" picks became "side" ones, the copies mo_similar refuses doubled and the field still check got 5 of 60 checks with over 0.5 % object pixels fit better by a variant (0 before).

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-20 19:44:10 +03:00
14ccef9a80 AMD FSR 3.1 upscales the 3D scene on the FidelityFX SDK 1.1.4 sources behind scene_upscaler = fsr3, off by default
scene_upscaler (off | dlss | fsr3, read at start) replaces the dlaa toggle on the Screen tab; without the key dlaa = 1
still means DLSS, so nobody gets FSR 3.1 without picking it. The DLSS mode row is the upscaling mode of both engines,
the DLSS model, characters' mask, fixed exposure and effects-after rows show only with DLSS, and the performance
overlay names the upscaler that ran and its mode. resolution.c sizes the scene for the engine picked at start.

runtime/vk_fsr3.cpp drives the SDK's upscaler context on its Vulkan backend directly, as Savage Wind's fsr3.cpp does:
one context per output size with the output as its largest render size (the modes and the dynamic resolution need
no new context), HDR input (the light target before the bloom, where DLSS reads it), render-size motion vectors with
the per-object pass, the camera jitter, the camera's planes and field of view from the scene block (vk_fg_camera; a
cut starts a new history), FSR's sharpening off. vk_dlaa.c calls it at the DLSS point after the recorder's drain;
the NGX path is left as it was. The backend's per-frame rings hold four dispatches: a dispatch waits for the one three
back should it still be in flight. Frame generation stays off while FSR 3.1 runs (logged once).

The SDK subset gains the upscaler component, its shader headers, FSR 1's (its RCAS) and its GLSL, from Savage Wind's
copy unchanged; tools/ffx_shaders.py --effects fu builds its blobs (HDR, 32- and 16-bit floats) into
spirv/ffx_fsr3upscaler_vk.h, and vk_fg.cpp's blob lookup hands those out through vk_fsr3_blob. The luma history's
storage image is declared rgba16f, its images' format (rgba8 in the SDK; the validation layer reported it).

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Revivedx <82332536+Revivedx@users.noreply.github.com>
2026-11-20 14:39:57 +03:00
Valiantsin DzerakhandClaude Opus 5.5 380e6ff30e Frame generation gets the per-object motion vectors DLSS uses, so Link running no longer shows a soft second outline and a trail in the generated pictures
The motion image FG read held the camera's motion alone (fg_motion called mv_pass without its object pass). The camera
follows Link, so his pixels carried the world's motion and the interpolation drew him twice, with a smeared strip where
he had been; NPCs and enemies did the same. fg_motion now runs mv_pass with the object pass (vk_mvobj_execute), and its
gates follow vk_objmv_wanted: DLSS once NGX came up (vk_dlaa_available, as before) or FG following the scene with its
objects (vk_fg_objects_on). vk_mvprev_frame keeps last frame's block tables for either, vk_usnap_begin makes the host
copy for either and vk_usnap_bytes reads the copy that submission wrote (FG turns on and off live), and
vk_dlaa_prev_camera answers whenever the scene is followed. FG is off with DLSS, so every DLSS gate evaluates as before.
fg_objects = 0 keeps the camera's motion alone for comparisons.

Measured on A (RX 9070, 7900X on 4 cores, 1080p, field, 60 s): 85.6 drawn with fg_objects = 0, 80.0 with 1 (about
0.8 ms a drawn frame on the replay thread, half of it the snapshot's host copy), object pass 0.035 ms GPU. Still grass
gets no false motion (SDK debug view: the game vector field stays within 0.2 px), and grass does not break up while the
camera turns. Validation: 0 errors.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-20 14:30:51 +03:00
Valiantsin DzerakhandClaude Opus 5.5 13b602f28f Keep #594's DLSS work off for players without DLSS
The scene patch, the heap sized for the largest DLSS scene and the function-entry hook that catches the game's render
size setup were on for every 1.4.3 player; they are on now only when DLSS can run (dlaa = 1, nvngx_dlss.dll there),
and the hook gives up after 1800 presents if the setup is never seen. Everyone else keeps the game's own scene size
and dynamic resolution, as before #594.

The uniform snapshot's second copy in host memory and the per-object motion's last-frame blocks are made only while
DLSS is available: without it nothing reads them, and the snapshot goes into the device copy in one pass as before.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-03 19:39:49 +03:00
DizzyMan1337andClaude Opus 5.5 ad3ecd8f3c Per-object motion vectors for DLSS, camera motion in double precision, Ultra Performance and DLSS presets.
Solid scene draws whose position reads an object's own uniform blocks (bones c5, place c7, c6) are drawn a second time
with a vertex shader variant (translator patch 0012) that computes last frame's position from last frame's blocks; the
game writes those into two copies that take turns, and the uniform snapshot keeps both (vk_mvprev.c). The draws are
listed during the replay and recorded through the recorder thread with repeated state skipped. Blocks whose last-frame
copies measured wrong (c9 and c11 of walls and cliffs) keep this frame's; positions without an object block (terrain,
grass) keep the camera's motion, which measured as good.

The camera motion is reprojected in double precision on the CPU: the float inverse made noise on far surfaces and a
still camera that DLSS followed into a wobble. DLSS gets Ultra Performance (a third of the output, also in the F1 menu)
and NVIDIA's recommended preset per mode (K, M for Performance, L for Ultra Performance; TOTK_DLSS_PRESET overrides).

Test switches: TOTK_MV_CHECK measures the motion vectors and DLSS's output against the picture, TOTK_DLAA_TIMES times
DLSS and the motion passes on the GPU. A reactive mask pass for transparents is included but off (TOTK_DLSS_REACTIVE):
its mask covers sky and fog too.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-03 17:31:58 +08:00