i cannot find any difference between the three possibilities of rotation in transform tab.
After i set there different numbers, which come in addition to zero, the pic looks exactly as before.
Are they really neccesary or am i overlooking sth?
Not a dev or very knowledgeable but the way I see and use it is.
First one is simple rotate most often used by drawing the line on a horizon or vertical
Second one is calculated from analyzing the geometry of the scene and part of a more complex set of transforms. This value attempts to set a geometrically “correct” but potentially not aesthetically great transform.
A set of tools designed to adjust the “geometrically correct” values to a visually appealing ones without altering the analysis of the scene.
So 1. and 2,3 are different tools but both can achieve rotation. 2 and 3 are different stages of a workflow. 2 setting correct values 3 allowing adjustment.
I can only confirm what @nosle wrote, as it is spot-on. The second module is under ‘Perspective’, which can be much more than simply rotating an image.
Thanks for your answers. I agree, that perspective can do much more, than just rotate.
For no. 1 the drawing line is a different handling than for 2 or 3 but in final result I come to the same result…
Still I cannot find the difference between those rotatings, I will investigate and learn…
They all rotate which can only be one thing. Rotate. They do however do so under different tools and for different aspects of the workflow. No use looking for how they are different in what they do because they are the same.
The difference comes from the placement in the processing pipeline. If the only adjustments you have are rotation, of course they have the same effect.
For the rotate and perspective sub modules, the order in the pipeline matches the order they appear in the UI. If you make any adjustments between the rotates, you will see they behave differently. For example, start with the values in the second screenshot. Now add some vertical correction. Take a snapshot of the result. Reset the rotations to zero and take another snapshot. Compare the two snapshots. They are different even though you might have thought the rotations cancel each other out.
Here’s another example. Start with the default values. Add some horizontal or vertical correction. Try each rotation slider and observe how they affect the result. The first two should behave the same because there are no adjustments in between them. The last one is different and behaves more like what you would expect rotation to look like because there are no more transformations applied after it.
@Lawrence37 thanks for your detailled explanation. I followed your instructions and see the difference now.
your explanation is worth to be brought to rawpedia