View Issue Details
| Category | |||||
|---|---|---|---|---|---|
| SSPBT：本体(SSP) | |||||
| Severity | minor | Reproducibility | always | ||
| Status | assigned | ||||
| Fixed in Version | 2.8 (リリース) | ||||
| Summary | 0000797: Turning off the high quality shell scaling option causes slowdown for some ghosts | ||||
| Description | Affects this ghost in particular: https://github.com/Zichqec/flakes_of_fancy I usually keep the high quality shell scaling option off, because I find that ghosts drawn in pixel art styles look better to me as long as the scaling is kept to values that divide cleanly (50%, 200%, etc). When the above ghost released, I was using SSP 2.8.32, and I could scale the shell down to 50% and the ghost would still run at normal speed. Now, on SSP 2.8.56, the ghost runs well at 100% shell scaling, but at 50% shell scaling it slows down to a crawl. If I turn on the high quality shell scaling option, the ghost runs at normal speed again. SSP/2.8.56 (20260715-12; Windows NT 10.0.26100) | ||||
| Attached Files | |||||
|
|
2.8.57 |
|
|
This issue is still occurring, I don't notice any change from the previous version. |
|
|
Does this degradation occur only at scales smaller than 100% (such as 50%)? Does it not occur at 200% or other scales larger than 100%? |
|
|
Could not reproduce on a different machine. Windows 11 Pro 25H2 (26200), 16 logical cores, display scale 100% (also tested 125%), flakes_of_fancy v1.0.0 at blizzard setting. Measured ssp.exe CPU time over 10s samples (% of one core): | Version | Display DPI | Shell scale | HQ scaling | CPU | |---------|-------------|-------------|------------|------| | 2.8.57 | 100% | 100% | off | 190% | | 2.8.57 | 100% | 50% | on | 171% | | 2.8.57 | 100% | 50% | off | 124% | | 2.8.57 | 125% | 50% | on | 200% | | 2.8.57 | 125% | 50% | off | 118% | | 2.8.32 | 100% | 100% | off | 140% | | 2.8.32 | 100% | 50% | off | 190% | | 2.8.32 | 100% | 50% | on | 175% | On this machine the reported condition (50%, HQ off) actually got faster from 2.8.32 to 2.8.57, and HQ-off was never slower than HQ-on on 2.8.5x. Numbers fluctuate somewht the gaps are well outside noise. To narrow this down, could the original reporter share: (1) ssp.exe CPU % in Task Manager while the slowdown is happening (high = CPU-bound, low = waiting on something)display scale (e.g. 100/125/150%), (4) snowintensity setting used, (5) single or multi monitor, HDR on/off, and whether this is a VM or Remote Desktop session. |
|
|
I rewrote the scaling logic from scratch, so there should be no environment-specific discrepancies, and it is likely faster now. 2.8.58,2.8.59 |
|
|
The issue seems to be resolved when high-quality shell scaling is turned off! I was trying all the different options to be sure, and oddly I noticed that the high-quality scaling option is fine when set to sizes lower than 50%, but at sizes between 100% and 50%, it struggles like before. 80% in particular is very slow. (Some of the sizes larger than 100% also seem a little slow, but that seems reasonable to me... this ghost is very intense to run in general.) For what it's worth, I'm replicating this on two computers. Laptop: SSP/2.8.59 (20260725-30; Windows NT 10.0.26200) Desktop (multi-display): SSP/2.8.59 (20260725-30; Windows NT 10.0.26100) The issues have been the same across both of them throughout. |
| Date Modified | Username | Field | Change |
|---|---|---|---|
| 2026-07-21 04:23 | guest | New Issue | |
| 2026-07-21 04:23 | guest | File Added: 26-7-20_02.gif | |
| 2026-07-23 09:25 | ponapalt | Assigned To | => ponapalt |
| 2026-07-23 09:25 | ponapalt | Status | new => closed |
| 2026-07-23 09:25 | ponapalt | Resolution | open => fixed |
| 2026-07-23 09:25 | ponapalt | Fixed in Version | => 2.8 (リリース) |
| 2026-07-23 09:25 | ponapalt | Note Added: 0001819 | |
| 2026-07-23 18:49 | guest | Status | closed => feedback |
| 2026-07-23 18:49 | guest | Resolution | fixed => reopened |
| 2026-07-23 18:49 | guest | Note Added: 0001820 | |
| 2026-07-23 20:25 | ponapalt | Note Added: 0001821 | |
| 2026-07-23 20:54 | guest | Note Added: 0001822 | |
| 2026-07-23 20:54 | guest | Status | feedback => assigned |
| 2026-07-25 21:06 | ponapalt | Note Added: 0001823 | |
| 2026-07-25 21:07 | ponapalt | Status | assigned => feedback |
| 2026-07-26 02:53 | guest | Note Added: 0001824 | |
| 2026-07-26 02:53 | guest | Status | feedback => assigned |
