Make fast and gaussian blur faster
#2Comparing
perf/faster-blur
(e86529a
) withcodspeed-wizard-1772277085312
(caea9d4
)×5.6
- Improvement
- 5
- Untouched
- 93
Benchmarks
98 total
fast blur: sigma 3.0
benches/blur.rs::benches::bench_fast_blur×5.6
254.3 ms45.3 ms
fast blur: sigma 7.0
benches/blur.rs::benches::bench_fast_blur×5.6
254.8 ms45.6 ms
fast blur: sigma 50.0
benches/blur.rs::benches::bench_fast_blur×5.2
260.6 ms50.3 ms
fast_blur
benches/fast_blur.rs::benches::bench_fast_blur×5.2
260.6 ms50.3 ms
gaussian blur: sigma 3.0
benches/blur.rs::benches::bench_fast_blur+17%
159.6 ms136.9 ms
gaussian blur: sigma 7.0
benches/blur.rs::benches::bench_fast_blur+9%
269.7 ms246.9 ms
gaussian blur: sigma 50.0
benches/blur.rs::benches::bench_fast_blur+2%
1.4 s1.4 s
zero-L8-file[100]
benches/encode.rs::benches::encode_all::encode-bmp+1%
16.2 µs16.1 µs
zero-L8-bufvec[100]
benches/encode.rs::benches::encode_all::encode-bmp0%
21.6 µs21.5 µs
zero-L8-file[200]
benches/encode.rs::benches::encode_all::encode-bmp0%
31.9 µs31.8 µs
sample_1.gif
benches/decode.rs::benches::load_all::load-Gif0%
26.7 µs26.6 µs
zero-L8-bufvec[200]
benches/encode.rs::benches::encode_all::encode-bmp0%
52.1 µs52 µs
zero-Rgb8-bufvec[64]
benches/encode.rs::benches::encode_all::encode-jpeg0%
152.1 µs151.9 µs
zero-L8-file[400]
benches/encode.rs::benches::encode_all::encode-bmp0%
96.7 µs96.6 µs
zero-L8-bufvec[64]
benches/encode.rs::benches::encode_all::encode-jpeg0%
113.7 µs113.6 µs
zero-L8-file[64]
benches/encode.rs::benches::encode_all::encode-jpeg0%
113.7 µs113.6 µs
zero-Rgb8-file[64]
benches/encode.rs::benches::encode_all::encode-jpeg0%
151.9 µs151.8 µs
zero-Rgb8-bufvec[100]
benches/encode.rs::benches::encode_all::encode-bmp0%
137.8 µs137.8 µs
test.jpg
benches/decode.rs::benches::load_all::load-Jpeg0%
160.6 µs160.5 µs
zero-Rgba8-rawvec[100]
benches/encode.rs::benches::encode_all::encode-bmp0%
106.1 µs106.1 µs
zero-L8-bufvec[400]
benches/encode.rs::benches::encode_all::encode-bmp0%
177.1 µs177 µs
zero-Rgba8-bufvec[100]
benches/encode.rs::benches::encode_all::encode-bmp0%
120 µs120 µs
zero-Rgb8-file[100]
benches/encode.rs::benches::encode_all::encode-bmp0%
123.1 µs123 µs
zero-Rgb8-bufvec[128]
benches/encode.rs::benches::encode_all::encode-jpeg0%
494.9 µs494.8 µs
zero-Rgba8-file[100]
benches/encode.rs::benches::encode_all::encode-bmp0%
100.1 µs100.1 µs
Commits
Click on a commit to change the comparison rangeBase
codspeed-wizard-1772277085312
caea9d4
×5.6
Use platform-aware fast rounding for FloatNearest::to_u8/to_u16
e86529a
14 hours ago
by art049