[Bug]: Incorect scaling of JPK image files

[derollins]

Describe the bug

When loading JPK file the raw data needs to be scaled and offset, the values of these offsets are found in the tags of the channel pages but there are multiple scaling factors currently AFMReader can choose the incorrect scaling factor leading to incorrect z values.

I have fixed this by seperating out the different scaling factors into their slots and selecting the 'default' slot that was used within the instrument.

Copy of the output

There is no error, the z values of loaded images are just incorrect.

To Reproduce

No response

AFMReader Version

0.1.0

Python Version

3.11

Operating System

Windows

Python Packages

No response

[derollins]

Some extra information.

The image below shows the output of 9 channels from the sample jpk used in examples and testing. When compared with other software that reads .jpk files (gwyddion, JPK DP, NanoLocz) the scaling is incorrect for each channel other than for measuredHeight.

Additionally phase can not be processed due to a key error when reading the scaling type. That gives this error message:

  File "C:\Users\ggjh246\OneDrive - University of Leeds\Code\test_jpk_load.py", line 15, in <module>
    raw_image, px2nm = load_jpk(file_path, channel)
                       ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "C:\Users\ggjh246\AppData\Local\anaconda3\envs\AFMReader_pip\Lib\site-packages\AFMReader\jpk.py", line 95, in load_jpk
    scaling_type = channel_page.tags["33027"].value
                   ~~~~~~~~~~~~~~~~~^^^^^^^^^
  File "C:\Users\ggjh246\AppData\Local\anaconda3\envs\AFMReader_pip\Lib\site-packages\tifffile\tifffile.py", line 11915, in __getitem__
    raise KeyError(key)
KeyError: '33027' 

Through introducing a function that can accurately read the channel and scaling data this can be corrected and the output matches the values from the other AFM image processing programmes.

[ns-rse]

Hi @derollins thanks for letting us know about this.

Reading through it sounds as though you've worked out how to solve this.

If so would you be able to make a Pull Request so we could include it please? To do so you'd have to make a fork of the
repository, clone that, make the change and then push back to your Fork, then you can make a pull request to
contribute to up-stream.

Apologies if I've got the wrong end of the stick and if you want any more help on how to do this please let me know.

[derollins]

No that's the help I was hoping for @ns-rse . I'll fork the repository with my fix and make a pull request.

[ns-rse]

Brilliant, thanks @derollins , shout if you have any problems on that side of things. 👍