gfxdisp / hdrutils Goto Github PK

When I pass demosaic_first=True to merge, the resulting values in HDR images are 4x larger than when demosaic_first=False. I assume the values should be comparable.

I get this error on Windows:

Traceback (most recent call last):
  File "C:\Users\manti\Documents\projects\lf_capture\pipeline\main.py", line 121, in <module>
    merge_HDR_all(args.scene, args.sigma, args.beta, args.color_calibration,
  File "C:\Users\manti\Documents\projects\lf_capture\pipeline\merge.py", line 96, in merge_HDR_all
    HDRutils.imwrite(out_file, merged)
  File "C:\Users\manti\anaconda3\envs\lf_capture\lib\site-packages\HDRutils\io.py", line 79, in imwrite
    elif img.dtype in (np.float32, np.float64, np.float128):
  File "C:\Users\manti\anaconda3\envs\lf_capture\lib\site-packages\numpy\__init__.py", line 303, in __getattr__
    raise AttributeError("module {!r} has no attribute "
AttributeError: module 'numpy' has no attribute 'float128'

I do not think that anyone would use float128 for pixel values, so I suggest removing this data type.

When passing 3-element vector wb to

merge(files, ... wb=wb):

I get a warning message:

.../HDRutils/merge.py:77: FutureWarning: elementwise comparison failed; returning scalar instead, but in the future will perform elementwise comparison
  if wb == 'camera':

The documentation could say in what format the numbers should be passed (numpy array?)

Hello! I encountered an issue while attempting to merge .CR2 RAW files for HDR processing. I noticed that the maximum pixel value in the resulting .exr file exceeds 10,000, and the average value appears to be quite high as well (around 130). I'm unsure if this is normal, considering the scene I photographed isn't particularly bright. Could there be an issue with my settings or configuration?

For your reference, I am using a Canon EOS 5D Mark III camera, and the code snippet I am using to merge the .CR2 files is as follows:
HDR_img = HDRutils.merge(CR2list, clip_highlights=True)[0]

Any insights or suggestions would be greatly appreciated. Thank you!

Hello! Can you provide examples of your input and output data that you showed in the paper as examples?

I need a newer imageio to import imageio.v3, but then I get:

ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts.
hdrutils 0.10 requires imageio==2.9.0, but you have imageio 2.21.2 which is incompatible.

It would make more sense to change all the requirements to >=