EwaldandClaude Sonnet 5
019a5c9172
Defining NODE_CLASS_MAPPINGS made ComfyUI skip comfy_entrypoint, so /imageops/viewmedia and node replacements were never registered. Load as a V3 extension only, move server.py to routes.py so it stops shadowing ComfyUI's server module, and transcode animated previews with PyAV instead of an external ffmpeg. - Require ComfyUI 0.19.0 (has_intermediate_output); CI floor updated. - Process 11 nodes on ComfyUI's torch device when the frames fit in free memory (IMAGEOPS_COMPUTE_DEVICE=cpu disables it). - Add core CURVE input to Color Correct, BOUNDING_BOX output/input to Crop and Crop Stitch, LAYERS output to Comp (ComfyUI 0.31+), essentials_category, and Merge blend modes with golden parity. - Report progress through the public set_progress API and drop the hidden unique_id input. - Add ImageOps settings and a refresh-previews command. - Fix ImageOpsComp rotation and preview saving (undefined to_display_range) and masked Transform edge colors. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
62 lines
1.8 KiB
Python
62 lines
1.8 KiB
Python
import pytest
|
|
import torch
|
|
|
|
from nodes._helpers import _apply_color_adjust
|
|
|
|
class TestHDRSupport:
|
|
def test_apply_color_adjust_preserves_hdr(self):
|
|
# Create an HDR tensor [B, H, W, C]
|
|
hdr_image = torch.tensor([[[[2.0, 5.0, 10.0]]]], dtype=torch.float32)
|
|
|
|
# Apply identity color adjustment
|
|
adjusted = _apply_color_adjust(
|
|
hdr_image,
|
|
temperature=0.0,
|
|
tint=0.0,
|
|
hue=0.0,
|
|
brightness=0.0,
|
|
contrast=0.0,
|
|
saturation=0.0,
|
|
vibrance=0.0,
|
|
gamma=1.0,
|
|
)
|
|
|
|
# It should not be clamped to 1.0
|
|
assert adjusted.max().item() > 1.0
|
|
assert torch.allclose(adjusted, hdr_image, atol=1e-4)
|
|
|
|
def test_apply_color_adjust_input_space_linear_skips_srgb_curve(self):
|
|
# A scene-linear/HDR source should not be run through the sRGB OETF/EOTF
|
|
# when input_space="linear", or brightness would double-gamma the result.
|
|
linear_image = torch.tensor([[[[0.25, 0.5, 2.0]]]], dtype=torch.float32)
|
|
|
|
adjusted = _apply_color_adjust(
|
|
linear_image,
|
|
temperature=0.0,
|
|
tint=0.0,
|
|
hue=0.0,
|
|
brightness=20.0,
|
|
contrast=0.0,
|
|
saturation=0.0,
|
|
vibrance=0.0,
|
|
gamma=1.0,
|
|
input_space="linear",
|
|
)
|
|
|
|
expected = linear_image * 1.2
|
|
assert torch.allclose(adjusted, expected, atol=1e-4)
|
|
|
|
|
|
|
|
|
|
def test_tensor_to_pil_clamps_hdr_for_display_only():
|
|
from nodes._helpers import _tensor_to_pil
|
|
|
|
image = torch.tensor([[[[2.0, -1.0, 0.5], [0.0, 1.0, 0.25]]]])
|
|
|
|
pil = _tensor_to_pil(image)
|
|
|
|
assert pil.mode == "RGB"
|
|
assert list(pil.getdata()) == [(255, 0, 128), (0, 255, 64)]
|
|
assert float(image.max()) == 2.0
|