Skip to main content

bm3dvk2

Block-Matching and 3D filtering, using Vulkan compute.

Ported from VapourSynth-BM3D. BM3D (spatial) and V-BM3D (temporal), basic and final estimates are consolidated into a single function. Temporal results are aggregated internally, so no separate aggregation call is needed.

Parameters

bm3dvk2.BM3D(vnode clip[, vnode ref=None, str profile="fast", float[] sigma=[10, 10, 10], int radius, int block_size, int block_step, int group_size, int bm_range, int bm_step, int ps_num, int ps_range, int ps_step, float th_mse, float hard_thr, int matrix=2])
  • clip: Clip to process. Gray or RGB of 16/32 bit float is supported. RGB is converted to the opponent color space internally and converted back for output. 16 bit float input is processed with half precision arithmetic where the device supports it.

  • ref: Reference clip, of the same format, dimensions and length as clip. When omitted, the basic estimate (collaborative hard-thresholding) is produced. When specified, the final estimate (collaborative Wiener filtering) is produced, with ref (usually a basic estimate) guiding block-matching and the Wiener filter.

  • profile: Preset of the defaults below: "fast", "lc" (low complexity), "np" (normal), "high" or "vn" (very noisy).

  • sigma: Strength of denoising per plane, in 8 bit scale, valid range [0, +inf). Unassigned planes take the value of the previous one. For Gray input, 0 returns the input unprocessed.

  • radius: Temporal radius, valid range [0, 16]. 0 denoises spatially (BM3D), greater values denoise temporally over radius * 2 + 1 frames (V-BM3D). Defaults to the radius of the profile, and the defaults of the other parameters follow the selected mode.

  • block_size: Size of a block, valid range [1, 64].

  • block_step: Step between reference blocks, valid range [1, block_size].

  • group_size: Maximum number of blocks in a group, valid range [1, 256].

  • bm_range: Radius of the block-matching search window, valid range [1, +inf).

  • bm_step: Step between block-matching search positions, valid range [1, bm_range].

  • ps_num: Number of the best matches of a frame that center the predictive search in the next frame, valid range [1, group_size]. Temporal only.

  • ps_range: Radius of the predictive search window, valid range [1, +inf). Temporal only.

  • ps_step: Step between predictive search positions, valid range [1, ps_range]. Temporal only.

  • th_mse: Threshold of mean square error for block-matching, valid range [0, +inf). Defaults depend on the profile and sigma[0].

  • hard_thr: Threshold of hard-thresholding, valid range (0, +inf). Basic estimate only.

  • matrix: Matrix coefficients of Gray input, used to normalize sigma and th_mse: 0 (GBR), 1 (BT.709), 2 (unspecified, chosen by resolution), 4 (FCC), 5 (BT.470BG), 6 (SMPTE 170M), 7 (SMPTE 240M), 9 (BT.2020 NCL), 10 (BT.2020 CL) or 100 (opponent color space). RGB input always uses the opponent color space.

Profile defaults

Values are basic / final.

profile radius block_size block_step group_size bm_range ps_num ps_range hard_thr
radius = 0
fast 8 / 8 8 / 7 8 / 8 9 / 9 2.7
lc 8 / 8 6 / 5 16 / 16 9 / 9 2.7
np 8 / 8 4 / 3 16 / 32 16 / 16 2.7
high 8 / 8 3 / 2 16 / 32 16 / 16 2.7
vn 8 / 11 4 / 6 32 / 32 16 / 16 2.8
radius > 0
fast 1 8 / 8 8 / 7 8 / 8 7 / 7 2 / 2 4 / 5 2.7
lc 2 8 / 8 6 / 5 8 / 8 9 / 9 2 / 2 4 / 5 2.7
np 3 8 / 8 4 / 3 8 / 8 12 / 12 2 / 2 5 / 6 2.7
high 4 8 / 8 3 / 2 8 / 8 16 / 16 2 / 2 7 / 8 2.7
vn 4 8 / 11 4 / 6 16 / 16 12 / 12 2 / 2 5 / 6 2.8

bm_step and ps_step default to 1. th_mse defaults to sigma[0] * 80 + 400 / sigma[0] * 10 + 200, or sigma[0] * 150 + 1000 / sigma[0] * 40 + 400 for "vn".

Installation

pip install -U vapoursynth-bm3dvk2

Metadata

Release files for vapoursynth-bm3dvk2 1.0

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for vapoursynth-bm3dvk2 1.0
File Size Uploaded
vapoursynth_bm3dvk2-1.0.tar.gz 180.9 kB Details

Built distributions (wheels)

Table of built distributions (wheels) for vapoursynth-bm3dvk2 1.0
File
vapoursynth_bm3dvk2-1.0-py3-none-win_amd64.whl Python 3 none Windows x86-64 Details
vapoursynth_bm3dvk2-1.0-py3-none-musllinux_1_2_x86_64.whl Python 3 none Linux musl 1.2+ x86-64 Details
vapoursynth_bm3dvk2-1.0-py3-none-musllinux_1_2_aarch64.whl Python 3 none Linux musl 1.2+ ARM64 Details
vapoursynth_bm3dvk2-1.0-py3-none-manylinux_2_24_x86_64.manylinux_2_28_x86_64.whl Python 3 none Linux glibc 2.24+ x86-64, Linux glibc 2.28+ x86-64 Details
vapoursynth_bm3dvk2-1.0-py3-none-manylinux_2_24_aarch64.manylinux_2_28_aarch64.whl Python 3 none Linux glibc 2.28+ ARM64, Linux glibc 2.24+ ARM64 Details
vapoursynth_bm3dvk2-1.0-py3-none-macosx_11_0_arm64.whl Python 3 none macOS 11.0+ ARM64 Details
vapoursynth_bm3dvk2-1.0-py3-none-macosx_10_15_x86_64.whl Python 3 none macOS 10.15+ x86-64 Details

Total release size: 3.8 MB

Release files / vapoursynth_bm3dvk2-1.0.tar.gz

Download URL vapoursynth_bm3dvk2-1.0.tar.gz
Size 180.9 kB
Tags Source
SHA-256 checksum
How to use checksums
056f61ef24ebbf64b95b379405da9dddc91498c8d5c9be32bcf9c76832bca015
BLAKE2b-256 checksum
How to use checksums
5a5cea529ac7af86b60915a59f25024e070a143ff563ad5f85a5c57544a6616f
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Oct 9, 2026.

Transparency log

Release files / vapoursynth_bm3dvk2-1.0-py3-none-win_amd64.whl

Download URL vapoursynth_bm3dvk2-1.0-py3-none-win_amd64.whl
Size 319.5 kB
Tags Python 3 Windows x86-64
SHA-256 checksum
How to use checksums
d213fe02022d62b77fab412f90e48028bbbb2045066ae2438507498902a8c633
BLAKE2b-256 checksum
How to use checksums
ab77779faa420ac3b69d706a45fcc1dc5862f9ee8023dca9affc3772ad880b4e
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Oct 9, 2026.

Transparency log

Release files / vapoursynth_bm3dvk2-1.0-py3-none-musllinux_1_2_x86_64.whl

Download URL vapoursynth_bm3dvk2-1.0-py3-none-musllinux_1_2_x86_64.whl
Size 1.2 MB
Tags Linux musl 1.2+ x86-64 Python 3
SHA-256 checksum
How to use checksums
43d970b244105f0386375a88ba5d31ad984e34cb50243d90dd7ecd7738d83ff8
BLAKE2b-256 checksum
How to use checksums
ca40d2d276c909423edc227fa29f86aa08b0dd8414e8192f9d74c37c90274d6e
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Oct 9, 2026.

Transparency log

Release files / vapoursynth_bm3dvk2-1.0-py3-none-musllinux_1_2_aarch64.whl

Download URL vapoursynth_bm3dvk2-1.0-py3-none-musllinux_1_2_aarch64.whl
Size 1.2 MB
Tags Linux musl 1.2+ ARM64 Python 3
SHA-256 checksum
How to use checksums
a67c608ae20fe10836b6a91c69035325d7b667c51baf55e87d1046f3dcdb5d37
BLAKE2b-256 checksum
How to use checksums
d832c489f04bc67c03cab3984545f5527c042f9644fa12dda58b23777a2423b3
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Oct 9, 2026.

Transparency log

Release files / vapoursynth_bm3dvk2-1.0-py3-none-manylinux_2_24_x86_64.manylinux_2_28_x86_64.whl

Download URL vapoursynth_bm3dvk2-1.0-py3-none-manylinux_2_24_x86_64.manylinux_2_28_x86_64.whl
Size 216.4 kB
Tags Linux glibc 2.24+ x86-64 Linux glibc 2.28+ x86-64 Python 3
SHA-256 checksum
How to use checksums
0ee0caaa18c12795eb8ddbb8190e8770c7e9d0a0f2bb9ab1461216f976eb03f4
BLAKE2b-256 checksum
How to use checksums
393cd14095913709a98b707d91277f31bed1d9dff4742e3e7ce12d7a264dc8d5
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Oct 9, 2026.

Transparency log

Release files / vapoursynth_bm3dvk2-1.0-py3-none-manylinux_2_24_aarch64.manylinux_2_28_aarch64.whl

Download URL vapoursynth_bm3dvk2-1.0-py3-none-manylinux_2_24_aarch64.manylinux_2_28_aarch64.whl
Size 211.6 kB
Tags Linux glibc 2.24+ ARM64 Linux glibc 2.28+ ARM64 Python 3
SHA-256 checksum
How to use checksums
92ab938708c30ede22330d5f1160d0961b451655a4c0740506fdf66b57ad655e
BLAKE2b-256 checksum
How to use checksums
12f34824fc4bd21f2408a8b4d3b6bec82b20a5b98ed3cb0134857297ae0ac67e
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Oct 9, 2026.

Transparency log

Release files / vapoursynth_bm3dvk2-1.0-py3-none-macosx_11_0_arm64.whl

Download URL vapoursynth_bm3dvk2-1.0-py3-none-macosx_11_0_arm64.whl
Size 203.4 kB
Tags Python 3 macOS 11.0+ ARM64
SHA-256 checksum
How to use checksums
c1248099bdaee98c69f6ad31e502513c6cbdc41ec41c5ed963441fa6e6c9f468
BLAKE2b-256 checksum
How to use checksums
505b0990f123f222dc49e892299dea2bf8f740cdafc24255cb61fabebaf9610b
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Oct 9, 2026.

Transparency log

Release files / vapoursynth_bm3dvk2-1.0-py3-none-macosx_10_15_x86_64.whl

Download URL vapoursynth_bm3dvk2-1.0-py3-none-macosx_10_15_x86_64.whl
Size 198.8 kB
Tags Python 3 macOS 10.15+ x86-64
SHA-256 checksum
How to use checksums
96a5a64c8da11e03b8dfdc67057bcf3e513280206ea6266c82729fe198905d2b
BLAKE2b-256 checksum
How to use checksums
88d7d0a4a13c834d2818f0b6d02f4fd01fc532ae21ce16ad743ec76cfa4c6072
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Oct 9, 2026.

Transparency log

Release history Release notifications | RSS feed

This release

1.0 This release

8 release files

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page