Game: Hair and foliage render as sparkling artifacts with severe frame-rate drops on Radeon Pro 580 #1836

Open
opened 2026-11-20 04:16:07 +00:00 by BeanBot · 1 comment
Member

Reported on Discord

Longshot version and channel

Stable v3.0.0 (source 87cb36aa).

What happened

Hair and foliage appear corrupted as dense bright sparkling/dotted patterns (see screenshots), reportedly throughout the game. The reporter says this visual glitch has always occurred and now also affects performance. They report typically 20–30 FPS; the screenshot shows 10 FPS. The profile log records severe slow periods, including 17.7 FPS over 5.8 seconds and intervals below 5 FPS.

Where in the game / steps / frequency

Reported in all parts of the game; the supplied bug reports identify HyruleField_Child. The exact actions/scenes that reproduce it and frequency are not specified beyond “all parts of the game.”

Graphics card and driver

Radeon Pro 580, AMD proprietary driver 24.9.1 (AMD proprietary shader compiler).

CPU, RAM and operating system

Intel Core i7-7700K @ 4.20 GHz (8 threads), 16 GB RAM, Windows 11 build 26300.

Graphics settings

1920x1080 output, Docked, game's own upscaler, 50% render resolution (960x540), sharpness 0%, frame pacing on. Other settings are listed in the attached report.txt.

Performance measurements

The 2026-11-19 23:13 report has HOS_FPS_LOG=1, HOS_THREAD_TIMES=5, RX_VK_PASS_TIMING=1 and RX_VK_SLOW_LOG=50 enabled. Its longshot_log.txt includes the profiling data. The earlier 22:50 report's log does not have HOS_THREAD_TIMES measurements. Screenshot shows 10 FPS, 102 ms, and 1% low 177 ms.

Logs and screenshots

Attached from the 2026-11-20 Discord messages: report.txt, longshot_log.txt, longshot_log.1.txt and picture.png from the later bug report; report.txt, longshot_log.txt, longshot_log.1.txt and picture.png from the earlier report. The screenshots show green sparkling corruption around Link and the surrounding scene in a barn.

Missing information

Please investigate whether this is the same renderer defect as Longshot #1204 (RX 580 2048SP), but note this report is a Radeon Pro 580 and the report does not establish they share the same cause. Unknown: exact reproducible steps/scene, how often the corruption occurs beyond “all parts,” whether performance is consistently low in a specific scene, and whether the reporter has tested a current nightly. V-Sync and frame-generation settings are not listed.

Reported on [Discord](https://discord.com/channels/1569849200047427645/1573185861166702742/1573185861166702742) ### Longshot version and channel Stable v3.0.0 (source 87cb36aa). ### What happened Hair and foliage appear corrupted as dense bright sparkling/dotted patterns (see screenshots), reportedly throughout the game. The reporter says this visual glitch has always occurred and now also affects performance. They report typically 20–30 FPS; the screenshot shows 10 FPS. The profile log records severe slow periods, including 17.7 FPS over 5.8 seconds and intervals below 5 FPS. ### Where in the game / steps / frequency Reported in all parts of the game; the supplied bug reports identify HyruleField_Child. The exact actions/scenes that reproduce it and frequency are not specified beyond “all parts of the game.” ### Graphics card and driver Radeon Pro 580, AMD proprietary driver 24.9.1 (AMD proprietary shader compiler). ### CPU, RAM and operating system Intel Core i7-7700K @ 4.20 GHz (8 threads), 16 GB RAM, Windows 11 build 26300. ### Graphics settings 1920x1080 output, Docked, game's own upscaler, 50% render resolution (960x540), sharpness 0%, frame pacing on. Other settings are listed in the attached report.txt. ### Performance measurements The 2026-11-19 23:13 report has HOS_FPS_LOG=1, HOS_THREAD_TIMES=5, RX_VK_PASS_TIMING=1 and RX_VK_SLOW_LOG=50 enabled. Its longshot_log.txt includes the profiling data. The earlier 22:50 report's log does not have HOS_THREAD_TIMES measurements. Screenshot shows 10 FPS, 102 ms, and 1% low 177 ms. ### Logs and screenshots Attached from the 2026-11-20 Discord messages: report.txt, longshot_log.txt, longshot_log.1.txt and picture.png from the later bug report; report.txt, longshot_log.txt, longshot_log.1.txt and picture.png from the earlier report. The screenshots show green sparkling corruption around Link and the surrounding scene in a barn. ### Missing information Please investigate whether this is the same renderer defect as Longshot #1204 (RX 580 2048SP), but note this report is a Radeon Pro 580 and the report does not establish they share the same cause. Unknown: exact reproducible steps/scene, how often the corruption occurs beyond “all parts,” whether performance is consistently low in a specific scene, and whether the reporter has tested a current nightly. V-Sync and frame-generation settings are not listed.
Author
Member
Files from the Discord report: - [report.txt](https://git.magicbean.dev/attachments/5589656d-32ab-4773-b7ac-03a436e324bd) - [longshot_log.txt](https://git.magicbean.dev/attachments/1528daaf-30a4-4f14-847c-eb7d329b62d0) - [longshot_log.1.txt](https://git.magicbean.dev/attachments/2ea14f33-3cff-4390-85c0-efa9dd991ead) - [report.txt](https://git.magicbean.dev/attachments/45f41c0d-625c-4993-b86e-3160e701170c) - [longshot_log.txt](https://git.magicbean.dev/attachments/2bfca4e2-e99e-4aee-823d-72c7aea0d601) - [longshot_log.1.txt](https://git.magicbean.dev/attachments/e86fdf06-7ff2-43d2-9879-366cb7bd7508)
Sign in to join this conversation.
No milestone
No project
No assignees
1 participant
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#1836
No description provided.