Nabu issueshttps://gitlab.esrf.fr/tomotools/nabu/-/issues2021-04-09T12:07:28+02:00https://gitlab.esrf.fr/tomotools/nabu/-/issues/235Adapt double flatfield for "Full Radios processing pipeline"2021-04-09T12:07:28+02:00Pierre PaleoAdapt double flatfield for "Full Radios processing pipeline"When using `FullRadiosPipeline`, we load `n` out of `n_angles` full radios, where `n < n_angles` in general for memory reasons.
However, double flatfield (**DFF**) need to load all the radios.
There are several solutions:
- Leave t...When using `FullRadiosPipeline`, we load `n` out of `n_angles` full radios, where `n < n_angles` in general for memory reasons.
However, double flatfield (**DFF**) need to load all the radios.
There are several solutions:
- Leave the things as they are: an "approximate DFF" is computed by doing the mean over `n` images (instead of `n_angles`). It works well on "bamboo" with `n = n_angles /2`.
- Perform the DFF step on sinograms. This certainly results in a different DFF.
- Create a "pre-process" class responsible for doing the DFF from `DataUrl`, computing the average frame by framehttps://gitlab.esrf.fr/tomotools/nabu/-/issues/155Alignment: focus calibration is missing some of the original features2020-08-27T11:07:00+02:00Nicola ViganoAlignment: focus calibration is missing some of the original featuresThe [focus.m](https://gitlab.esrf.fr/tomotools/octave_archive/-/blob/master/m/tomotools/focus.m) function presents more options than the simple 'std' merit function.
These other options (at least the power spectral density method) need ...The [focus.m](https://gitlab.esrf.fr/tomotools/octave_archive/-/blob/master/m/tomotools/focus.m) function presents more options than the simple 'std' merit function.
These other options (at least the power spectral density method) need to be ported. This is not a critical issue, so no due date or milestone is expected.Nicola ViganoNicola Viganohttps://gitlab.esrf.fr/tomotools/nabu/-/issues/318CTF: use different factor for padding2022-09-09T14:14:40+02:00Pierre PaleoCTF: use different factor for paddingIn CTF, when `padded_shape="auto"`, each dimension is only doubled:
```python
# [...]
elif padded_shape == "auto":
padded_shape = (2 * self.shape[0], 2 * self.shape[1])
# [...]
```
It would be good to...In CTF, when `padded_shape="auto"`, each dimension is only doubled:
```python
# [...]
elif padded_shape == "auto":
padded_shape = (2 * self.shape[0], 2 * self.shape[1])
# [...]
```
It would be good to use `get_next_power()` as done for `SinoFilter`.