nodes/CosmicLaca-ComfyUI_Primere_Nodes
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

V 2.0.0 - Rasterix

Browse Source
This commit is contained in:
DESKTOP-TVBJISQ\Primere
2026-03-18 17:37:52 +01:00
parent 24d5d65232
commit 74a9dd92d3
4 changed files with 133 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files
Nodes/Dashboard.py
+76 -1
View File
@@ -68,6 +68,12 @@ from transformers import AutoTokenizer, T5Tokenizer, T5EncoderModel, AutoModelFo
from ..components.sana.diffusion.data.datasets.utils import ASPECT_RATIO_512_TEST, ASPECT_RATIO_1024_TEST, ASPECT_RATIO_2048_TEST
import node_helpers
from comfy_api.latest import ComfyExtension, io
from ..components.images import img_shade_level as img_shade_level
from ..components.images import img_brightness_contrast as img_brightness_contrast
from ..components.images import img_color_balance as img_color_balance
from ..components.images import img_hue_saturation as img_hue_saturation
from ..components.images import img_levels_auto as img_levels_auto
from ..components.images import isgen_detect_ext_full as isgen_detect_ext_full
class PrimereSamplersSteps:
CATEGORY = TREE_DASHBOARD
@@ -2139,4 +2145,73 @@ class PrimereUpscaleModel:
def load_upscaler(self, model_name):
out = nodes_upscale_model.UpscaleModelLoader.execute(model_name)[0]
return (out, model_name,)
return (out, model_name,)
class PrimereRasterix:
RETURN_TYPES = ("IMAGE",)
RETURN_NAMES = ("IMAGE",)
FUNCTION = "primere_rasterix"
CATEGORY = TREE_DASHBOARD
@classmethod
def INPUT_TYPES(cls):
return {
"required": {
"image": ("IMAGE", {"forceInput": True}),
"auto_normalize": ("BOOLEAN", {"default": False, "label_off": "No auto levels", "label_on": "Apply auto levels"}),
"auto_levels_threshold": ("FLOAT", {"default": 2.0, "min": 0.0, "max": 10.0, "step": 0.1}),
"shade_level": ("FLOAT", {"default": 0, "min": -100, "max": 100, "step": 1}),
"shade_radius": ("FLOAT", {"default": 0, "min": 0, "max": 50, "step": 0.5}),
"brightness": ("FLOAT", {"default": 0, "min": -150, "max": 150, "step": 1}),
"contrast": ("FLOAT", {"default": 0, "min": -50, "max": 100, "step": 1}),
"use_legacy": ("BOOLEAN", {"default": False, "label_off": "Use non-linear shift", "label_on": "Use adaptive offset"}),
"color_balance_cyan_red": ("FLOAT", {"default": 0, "min": -100, "max": 100, "step": 1}),
"color_balance_magenta_green": ("FLOAT", {"default": 0, "min": -100, "max": 100, "step": 1}),
"color_balance_yellow_blue": ("FLOAT", {"default": 0, "min": -100, "max": 100, "step": 1}),
"color_balance_tone": (["highlights", "midtones", "shadows"], {"default": "midtones"}),
"color_balance_preserve_luminosity": ("BOOLEAN", {"default": False, "label_off": "Modify luminosity", "label_on": "Restore original luminosity"}),
"hue_saturation_channel": (["master", "r", "g", "b"], {"default": "master"}),
"hue_saturation_hue": ("FLOAT", {"default": 0, "min": -180, "max": 180, "step": 1}),
"hue_saturation_saturation": ("FLOAT", {"default": 0, "min": -100, "max": 100, "step": 1}),
"hue_saturation_lightness": ("FLOAT", {"default": 0, "min": -100, "max": 100, "step": 1}),
"hue_saturation_vibrance": ("FLOAT", {"default": 0, "min": -100, "max": 100, "step": 1}),
"ai_detection": ("BOOLEAN", {"default": False, "label_off": "AI detection bypass off", "label_on": "AI detection bypass on"}),
"grain_intensity": ("FLOAT", {"default": 6.5, "min": 0.0, "max": 20.0, "step": 0.5}),
"freq_strength": ("FLOAT", {"default": 0.019, "min": 0.0, "max": 0.1, "step": 0.001}),
"variance_strength": ("FLOAT", {"default": 0.32, "min": 0.0, "max": 1.0, "step": 0.01}),
"ca_strength": ("FLOAT", {"default": 1.2, "min": 0.0, "max": 5.0, "step": 0.1}),
"vignette_strength": ("FLOAT", {"default": 0.18, "min": 0.0, "max": 1.0, "step": 0.01}),
"unsharp_percent": ("INT", {"default": 38, "min": 0, "max": 150, "step": 1}),
"jpeg_quality": ("INT", {"default": 95, "min": 60, "max": 100, "step": 1}),
"jpeg_cycles": ("INT", {"default": 3, "min": 0, "max": 6, "step": 1}),
}
}
def primere_rasterix(self, image, auto_normalize, auto_levels_threshold, shade_level, shade_radius, brightness, contrast, use_legacy, color_balance_cyan_red, color_balance_magenta_green, color_balance_yellow_blue, color_balance_tone, color_balance_preserve_luminosity, hue_saturation_channel, hue_saturation_hue, hue_saturation_saturation, hue_saturation_lightness, hue_saturation_vibrance, ai_detection, grain_intensity, freq_strength, variance_strength, ca_strength, vignette_strength, unsharp_percent, jpeg_quality, jpeg_cycles):
pil_img = utility.tensor_to_image(image)
if auto_normalize:
pil_img = img_levels_auto.img_levels_auto(image=pil_img, auto_normalize=auto_normalize, threshold=auto_levels_threshold)
if brightness != 0 or contrast != 0:
pil_img = img_brightness_contrast.img_brightness_contrast(image=pil_img, brightness=brightness, contrast=contrast, use_legacy=use_legacy)
if color_balance_cyan_red != 0 or color_balance_magenta_green != 0 or color_balance_yellow_blue != 0:
pil_img = img_color_balance.img_color_balance(image=pil_img, cyan_red=color_balance_cyan_red, magenta_green=color_balance_magenta_green, yellow_blue=color_balance_yellow_blue, tone=color_balance_tone, preserve_luminosity=color_balance_preserve_luminosity)
if hue_saturation_hue != 0 or hue_saturation_saturation != 0 or hue_saturation_lightness != 0 or hue_saturation_vibrance != 0:
pil_img = img_hue_saturation.img_hue_saturation(image=pil_img, channel=hue_saturation_channel, hue=hue_saturation_hue, saturation=hue_saturation_saturation, lightness=hue_saturation_lightness, vibrance=hue_saturation_vibrance)
shade_radius = None if shade_radius == 0 else shade_radius
if shade_level != 0:
pil_img = img_shade_level.img_shade_level(image=pil_img, shade_level=shade_level, radius=shade_radius)
if ai_detection:
pil_img = detect_ext_full.bypass_ai_detector(image=pil_img, grain_intensity=grain_intensity, freq_strength=freq_strength, variance_strength=variance_strength, ca_strength=ca_strength, vignette_strength=vignette_strength, unsharp_percent=unsharp_percent, jpeg_quality=jpeg_quality, jpeg_cycles=jpeg_cycles)
return (utility.image_to_tensor(pil_img),)
__init__.py
+2
View File
@@ -58,6 +58,7 @@ NODE_CLASS_MAPPINGS = {
"PrimereNetworkTagLoader": Dashboard.PrimereNetworkTagLoader,
"PrimereModelKeyword": Dashboard.PrimereModelKeyword,
"PrimereUpscaleModel": Dashboard.PrimereUpscaleModel,
"PrimereRasterix": Dashboard.PrimereRasterix,
"PrimerePrompt": Inputs.PrimereDoublePrompt,
"PrimereStyleLoader": Inputs.PrimereStyleLoader,
@@ -137,6 +138,7 @@ NODE_DISPLAY_NAME_MAPPINGS = {
"PrimereModelKeyword": "Primere Model Keyword",
"PrimereConceptDataTuple": "Primere Concept Tuple",
"PrimereUpscaleModel": "Primere Upscale Models",
"PrimereRasterix": "Primere Rasterix (The ToneLab)",
"PrimerePrompt": "Primere Prompt",
"PrimereStyleLoader": "Primere Styles",
components/images/img_color_balance.py
+47
View File
@@ -0,0 +1,47 @@
import numpy as np
from PIL import Image
def img_color_balance(image: Image.Image, cyan_red: float = 0, magenta_green: float = 0, yellow_blue: float = 0, tone: str = 'midtones', preserve_luminosity: bool = True,) -> Image.Image:
VALID_TONES = {'shadows', 'midtones', 'highlights'}
tone = tone.strip().lower()
if tone not in VALID_TONES:
raise ValueError(f"tone must be one of {VALID_TONES}, got '{tone}'")
img = image.convert("RGB")
arr = np.array(img, dtype=np.float32)
lum = (0.299 * arr[:,:,0] + 0.587 * arr[:,:,1] + 0.114 * arr[:,:,2]) / 255.0
if tone == 'shadows':
mask = 1.0 - lum
mask = mask ** 1.5
elif tone == 'highlights':
mask = lum
mask = mask ** 1.5
else:
mask = 1.0 - np.abs(2.0 * lum - 1.0)
mask = mask ** 0.8
mask = mask[:, :, np.newaxis]
delta_r = cyan_red
delta_g = magenta_green
delta_b = yellow_blue
shift = np.array([delta_r, delta_g, delta_b], dtype=np.float32)
adjusted = arr + mask * shift
if preserve_luminosity:
lum_before = (0.299 * arr[:,:,0] + 0.587 * arr[:,:,1] + 0.114 * arr[:,:,2])
lum_after = (0.299 * adjusted[:,:,0] + 0.587 * adjusted[:,:,1] + 0.114 * adjusted[:,:,2])
with np.errstate(invalid="ignore", divide="ignore"):
scale = np.where(lum_after > 1e-6, lum_before / lum_after, 1.0)
scale = scale[:, :, np.newaxis]
scale_blended = 1.0 + mask[:,:,0:1] * (scale - 1.0)
adjusted = adjusted * scale_blended
result = np.clip(adjusted, 0, 255).astype(np.uint8)
return Image.fromarray(result, mode="RGB")
components/utility.py
+8
View File
@@ -1195,6 +1195,14 @@ def tensor_to_image(tensor):
image = Image.fromarray(image_np, mode='RGB')
return image
def image_to_tensor(image: Image.Image | None) -> torch.Tensor | None:
if image is None:
return None
rgb_image = image.convert("RGB")
image_array = np.array(rgb_image).astype(np.float32) / 255.0
return torch.from_numpy(image_array)[None,]
def florence_img2prompt(model, processor, image, max_new_tokens, num_beams, do_sample, text_input=None, llm_options=None):
if text_input is None:
text_input = 'detailed enhanced prompt for text2image models'